TENNESSEE ARCHAEOLOGY

Volume 8 Summer 2016 Numbers 1-2

EDITORIAL COORDINATORS Michael C. Moore Tennessee Division of Archaeology TENNESSEE ARCHAEOLOGY Kevin E. Smith Middle Tennessee State University VOLUME 8 Summer 2016 NUMBERS 1-2

EDITORIAL ADVISORY COMMITTEE 1 EDITORS CORNER Paul Avery Cultural Resource Analysts, Inc. 4 The Old Man and the Pleistocene: John Broster and Paleoindian Period Archaeology in the Mid-South – Jared Barrett Introduction to the Special Volume TRC Inc. D. SHANE MILLER AND JESSE W. TUNE

Sierra Bow 8 A Retrospective Peek at the Career of John Bertram University of Tennessee Broster Andrew Brown MICHAEL C. MOORE, KEVIN E. SMITH, AARON DETER- University of North Texas WOLF, AND DAVID E. STUART

Hannah Guidry 24 The Paleoindian and Early Archaic Record in TRC Inc. Tennessee: A Review of the Tennessee Fluted Point Survey Michaelyn Harle JESSE W. TUNE Tennessee Valley Authority 42 Quantifying Regional Variation in Terminal Phillip Hodge Pleistocene Assemblages from the Lower Tennessee Tennessee Department of Transportation River Valley Using Chert Sourcing RYAN M. PARISH AND ADAM FINN Shannon Hodge Middle Tennessee State University 59 A Preliminary Report on the Late Pleistocene and Sarah Levithol Early Holocene Archaeology of Rock Creek Mortar Tennessee Division of Archaeology Shelter, Upper Cumberland Plateau, Tennessee JAY FRANKLIN, MAUREEN HAYS, FRÉDÉRIC SURMELY, Ryan Parish ILARIA PATANIA, LUCINDA LANGSTON, AND TRAVIS BOW University of Memphis 78 Colonization after Clovis: Using the Ideal Free Tanya M. Peres Distribution to Interpret the Distribution of Late Florida State University Pleistocene and Early Holocene Archaeological Sites in the Duck River Valley, Tennessee Jesse Tune D. SHANE MILLER AND STEPHEN B. CARMODY Fort Lewis College 102 The Paleoindian and Early Archaic Hilltop Tennessee Archaeology is published Occupations at the Topper Site semi-annually in electronic print format DEREK T. ANDERSON, ASHLEY M. SMALLWOOD, ALBERT under the auspices of the Tennessee C. GOODYEAR, AND SARAH E. WALTERS Council for Professional Archaeology. 114 Clovis Blade Technology and Tool Use along the Correspondence about manuscripts for South Atlantic Coastal Plain and Piedmont of the the journal should be addressed to Lower Southeast Michael C. Moore, Tennessee Division of DOUGLAS SAIN AND ALBERT C. GOODYEAR Archaeology, Cole Building #3, 1216 Foster Avenue, Nashville TN 37243. 132 Making a Difference: John B. Broster and Paleoindian Archaeology in Tennessee The Tennessee Council for Professional DAVID G. ANDERSON Archaeology disclaims responsibility for statements, whether fact or of opinion, made by contributors.

On the Cover: John B. Broster and his Career Achievement Award presented by the Tennessee Council for Professional Archaeology, January 26, 2013 (Photograph, Kevin E. Smith). EDITORS CORNER

We are pleased to devote the eighth volume of Tennessee Archaeology to a special double issue guest-edited by Jesse W. Tune and D. Shane Miller honoring the contributions of John B. Broster (for more background on John, see particularly Miller and Tune; Moore et al.; David Anderson, this volume). As always, we appreciate the contributions of the authors and extend our thanks to the reviewers who help make this peer-reviewed e-journal possible. We report several items of note on activities in Tennessee archaeology since our last Editors Corner. Following the rediscovery of portions of the Mississippian-era salt manufacturing site in downtown Nashville in 2014, Dr. Rex Weeks, Museum Curator at the Tennessee State Museum coordinated the design and installation in August 2015 of a new permanent display about the site (see “Editors Corner,” Tennessee Archaeology 7(2):103-104). State museum staff also assisted in the design of a series of “prehistory markers” installed by Metro Nashville/Davidson County on the greenway near the new Nashville Sounds Stadium. We are pleased to see this important aspect of Nashville’s prehistory FIGURE 1. New permanent exhibition on the Mississippian salt highlighted with the new exhibit and industry at the Tennessee State Museum (Kevin E. Smith). signage.

FIGURE 2. Five new “Prehistory” markers on the greenway system near the Sounds Stadium (Rex Weeks)

FIGURE 3. “Salt Industry” sign on the greenway (Rex Weeks).

1 Tennessee Archaeology 8(1-2) Summer 2016

During November 2015, the Southeastern Archaeological Conference met for the second time in 72 years at the Doubletree by Hilton Downtown Nashville. Hosted by the Tennessee Division of Archaeology, Middle Tennessee State University, and the Tennessee Department of Transportation, SEAC Nashville 2015 represents the tenth meeting of the conference in Tennessee: Chattanooga (2001); Knoxville (1950, 1968, 1978, 1995, 2007); Memphis (1973, 1982); Nashville (1986, 2015). Two well-attended excursions visited Old Stone Fort State Archaeological Park in Manchester and Wynnewood State Historic Site and the Castalian Springs Mounds State Archaeological Area in Castalian Springs. Bulletin 58, containing FIGURE 4. SEAC Nashville the papers and abstracts of the 2015 Nashville conference, is 2015 Conference Bulletin archived by the Southeastern Archaeological Conference: http://www.southeasternarchaeology.org/publications/seac-bulletins/. A memorable conference event was the Thursday evening reception at the Tennessee State Museum attended by an estimated 500+ southeastern archaeologists. On display was the temporary exhibition “ANCESTORS: Ancient Native American Stone Sculptures of Tennessee” (31 Oct 2015-15 May 2016). ANCESTORS was the first major temporary exhibition of Tennessee’s prehistoric artifacts by the state museum in recent years. Curated by Rex Weeks of the Tennessee State Museum with assistance from guest curators Kevin E. Smith (Middle Tennessee State University) and Robert V. Sharp, the exhibition assembled 28 of the Tennessee-Cumberland style statuary – including 14 from the Smithsonian Institution, two from the Metropolitan Museum of Art in New York City, two from the McClung Museum of Natural FIGURE 5. SEAC Reception at the Tennessee State Museum (Phillip History and Culture, five from the state museum's Hodge) collections, and five that are held in private collections. A centerpiece of the exhibition was the assembly for the first time in the modern era of all four statues from Sellars Farm State Archaeological Area (40WI1). Sellars is managed by Long Hunter State Park and is open to the public during daylight hours for self-guided tours. The curators are planning to issue a post-exhibition catalog including updated information resulting from new and on-going research in the future. We also take this opportunity to recognize the passing of two valued contributors to Tennessee archaeology. We extend our FIGURE 6. Four statues from Sellars Farm State condolences to their family, friends, and Archaeological Area greeting visitors to the exhibition (Kevin E. Smith). colleagues. They will be missed.

2 Editors Corner

Robert Ashley (“Bob”) Pace (26 Oct 1953-4 Mar 2015) passed away in Charleston, South Carolina at the age of 61. Born in Knoxville, Tennessee to Blanche Davis Pace and Ashe Heath Pace, he earned a degree in anthropology from the University of Tennessee. Bob’s professional archaeological career spanned over 30 years, much of it spent working on projects in Tennessee. While in Knoxville, he participated in some of the early archaeological surveys during the 1980s development of Big South Fork National River and Recreation Area. Also early in his career, Bob worked for the Schiele Museum in Charlotte, North Carolina, before locating to Franklin, Tennessee where he was employed for many years by DuVall & Associates as a field archaeologist. While working for DuVall, he met Nancy Tinker and the two were married on October 19, 2003. Bob is remembered FIGURE 7. Bob Pace on a camping trip as “a gentle thoughtful man with an acerbic wit” at Big South Fork in 1983 (Jeffrey W. Gardner) (Obituary, Charleston Post and Courier, 10 Mar 2015). Robert Phillip (“Bobby”) Hulan of Ashland City, Tennessee, passed away on April 30, 2015 at the age of 62. Preceded in death by his parents Granville and Billy Z. Hulan, he is survived by his wife Janet and sister Cathi Brumley (Obituary, Nashville Tennessean, 6 May 2015). An avocational archaeologist, Bobby is best remembered for his many contributions to the Tennessee Paleoindian Projectile Point and Site Survey and particularly at the Widemeier site (40DV9) – cited directly in four of the articles included in this volume. FIGURE 8. Bobby Hulan (Photograph courtesy, Janet Hulan). Of Delaware (Native American) descent, Bobby attended Cheatham County High School and was subsequently a high-steel ironworker until a four-story fall in 1978 left him disabled: “The best ironworkers are Indians… they walk one foot in front of the other. This Indian fell… I’m doing o.k. now…. I have my own business called Broken Hand Art Works” (quoted in Howard West, 2011, Last Grand Adventure, published in eBook format by eBookit.com, accessed 1 Jun 2016). In Bobby’s own words, he was interested in “Native American Living History, all aspects of ancient man and the ability to live and survive. My grandfather taught me at an early age to live with the deep forest, not against it. I was instructed [in] the medicinal plants, herbs, and trees that my ancestors used. I still gather and apply, what I learned as a young child” (Bobby Hulan. (n.d.) LinkedIn [Profile page]. Retrieved 1 Jun 2016, from https://www.linkedin.com/in/bobby-hulan-bb2b8a57).

3 INTRODUCTION TO THE SPECIAL VOLUME

THE OLD MAN AND THE PLEISTOCENE: JOHN BROSTER AND PALEOINDIAN ARCHAEOLOGY IN THE MID-SOUTH

D. Shane Miller and Jesse W. Tune

This double issue of Tennessee of the Mid-South for a substantial portion of Archaeology is based on the symposium his career. Recent Research and Future Directions in One of the lasting contributions from Southeastern Paleoindian Archaeology John’s prolific career is that he set the tone held in honor of John Broster during the for how Paleoindian period research Southeastern Archaeological Conference should be conducted in Tennessee. First, annual meeting in Tampa, Florida in 2013 John (and his colleague Mark Norton) have (Figure 1). We were able to assemble a worked extensively with the public broad array of contributors that reflect the examining private collections, which led to impact that John has had at a state, a revitalization of the Tennessee Fluted regional, and national level in regards to Point Survey in the late 1980s and has Paleoindian period archaeology. It is our since expanded to include nearly 5,500 hope that these contributions display our artifacts (Broster and Norton 1992, 1996; gratitude for the way in which John has Tune, this volume). As a result of this conducted research, collaborated with outreach and the lasting relationships he’s colleagues, and helped dozens of students built with private collectors across the state get their research off the ground. of Tennessee, John was able to locate, Moreover, over the course of John’s career analyze, and publish information on he has helped address some of the “Big several now famous sites like Carson- Questions” in Paleoindian archaeology. Conn-Short, Coats-Hines, Widemeier, We would argue that his ability to connect Sinclair, and Puckett (Broster and Norton the archaeology of the Mid-South to 1996; Broster et al. 2013). By publishing broader research questions come from a information on these sites and isolated lifetime of conducting fieldwork in a variety finds through numerous journal articles of contexts – from Oaxaca to Sante Fe to and the Tennessee Fluted Point Survey (a Pinson Mounds – as Moore et al. (this version of which can be downloaded volume) discuss with a light dose of candid at http://pidba.utk.edu/tennessee.htm), humor. Upon reading their contribution for John provided avenues through which the first time, I felt as though John’s life was Tennessee’s rich Late Pleistocene and pulled from the pages of a novel by Ernest Early Holocene archaeological record Hemingway, perhaps like “Santiago” from could be available to a broader audience. The Old Man and the Sea (1952) – a The significance of this cannot be larger-than-life character who didn’t back overstated, as it has facilitated research down from an even larger challenge. and collaborative partnerships across However, in John’s case, his challenge North America. wasn’t reeling a 20-foot-long Marlin by The areas where John has focused hand. Instead, John went toe-to-toe with most of his attention is the lower the Pleistocene and Early Holocene record Tennessee and middle Cumberland river

4 The Old Man and the Pleistocene

fauna occurrences are rare to non-existent (Breitburg and Broster 1994; Broster et al. 2013; Lane and Anderson 2001; Maggard and Stacklebeck 2008). Consequently, we can now conclude that early sites are relatively rare at higher elevations, which makes sites like Rock Creek Mortar Shelter (Franklin et al., this volume) particularly unique and important, and allows us to begin formulating hypotheses about the spatial and temporal variability of the colonization process in the Mid-South and subsequent responses to climate change (Miller and Carmody, this volume). John Broster’s influence on FIGURE 1. Jesse Tune, John Broster, and Paleoindian archaeology also extends Shane Miller at the session in Tampa, beyond Tennessee. For example, Florida. researchers in other areas of eastern North valleys, in particular at sites like Carson- America use the record from Tennessee as Conn-Short (Broster and Norton 1993, an important baseline for comparison. For 1996; Nami et al. 1996; Norton and Broster example, Anderson et al. (this volume) and 2008; Stanford et al. 2006), Coats-Hines Sain and Goodyear (this volume) are two (Breitburg et al. 1996; Deter-Wolf et al. studies focused on the Lower Atlantic 2011), Johnson (Barker and Broster 1996; Coastal Plain, and the Savannah River Broster and Barker 1992), Kirk Point Valley in particular. Because buried, (McNutt et al. 2008), Nuckolls Extension stratified Late Pleistocene and Early (Norton and Broster 1992), Pierce (Broster Holocene sites are so rare in the 1982), Puckett (Norton and Broster 1993), southeastern United States, sites in middle Trull (Norton et al. 1998), Twelkemeier Tennessee like Carson-Conn-Short on the (Broster and Norton 1990), Widemeier Lower Tennessee River and the Johnson (Broster et al. 2006), and numerous private site on the Cumberland River provide collections (Broster et al. 2013). It’s largely some of the only data points to begin because of his efforts that we now know examining the regional variability in the that middle Tennessee has one of the early record, which can then be used to richest Late Pleistocene and Early evaluate models for the colonization of Holocene archaeological records in North North America (e.g., Smallwood 2012). America, both in terms of density and More broadly, Anderson (this volume) diversity. Parris and Finn (this volume) discusses in greater detail the lasting continue the tradition of public outreach impact of John Broster on the Late and collections-based research with their Pleistocene and Early Holocene record of analysis of the Jim Parris Collection in eastern North America. Finally, unlike Hardin County. In addition to Santiago from The Old Man and the Sea, demonstrating where the early record is John did not return to port empty-handed. more prevalent, John’s body of research Instead, all of the contributors to this has also helped us see where the early volume would like to congratulate John on archaeological sites and Late Pleistocene a long and fruitful career. He truly is a

5 Tennessee Archaeology 8(1-2) Summer 2016

larger-than-life figure in the archaeology of the Pleistocene 10:3-5. Tennessee. 1996 Recent Paleoindian Research in Tennessee. In The Paleoindian and Early Archaic Southeast, References edited by D. G. Anderson and K. E.

Sassaman, pp. 288-297. University Barker, Gary and John B. Broster of Alabama Press, Tuscaloosa. 1996 The Johnson Site (DV400): A

Dated Paleoindian and Early Broster, John B., Mark R. Norton, Bobby Archaic Occupation in Tennessee's Hulan, and Ellis Durham Central Basin. Journal of Alabama 2006 A Preliminary Analysis of Clovis Archaeology 42(2):97-153. through Early Archaic Components

at the Widemeier Site (40DV9), Breitburg, Emanuel, and John B. Broster Davidson County, Tennessee. 1994 Paleoindian Site, Lithic, and Tennessee Archaeology 2(2):120- Mastodon Distributions in 127. Tennessee. Current Research in

the Pleistocene 11:9-11. Broster, John B., Mark R. Norton, D.

Shane Miller, Jesse W. Tune and Breitburg, Emanuel, John B. Broster, A.L. Jonathan Baker Reesman, and R.G. Stearns 2013 Tennessee's Paleoindian Record: 1996 The Coats-Hines Site: Tennessee's The Cumberland and Lower First Paleoindian-Mastodon Tennessee River Watersheds. In In Association. Current Research in the Eastern Fluted Point Tradition, the Pleistocene 13:6-8. edited by J. A. M. Gingerich, pp.

299-314. University of Utah Press, Broster, John B. Salt Lake City. 1982 Paleo-Indian Habitation at the

Pierce Site (40CS24); Chester Deter-Wolf, Aaron, Jesse W. Tune and County, Tennessee. Tennessee John B. Broster Anthropologist 7:93-104. 2011 Excavations and Dating of Late

Pleistocene and Paleoindian Broster, John B., and Gary L. Barker Deposits at the Coats-Hines Site, 1992 Second Report of Investigations at Williamson County, Tennessee. the Johnson Site (40DV400): The Tennessee Archaeology 5(2):142- 1991 Field Season. Tennessee 156. Anthropologist 17(2):120-130.

Hemingway, Charles Broster, John B. and Mark R. Norton 1952 The Old Man and the Sea. Charles 1990 Lithic Analysis and Paleo-Indian Scribner and Sons, New York. Utilization of the Twelkemeier Site

(40HS173). Tennessee Lane, Leon and David G. Anderson Anthropologist 15:115-131. 2001 Paleoindian Occupations of the 1993 The Carson-Conn-Short Site Southern Appalachians. In (40BN190): An Extensive Clovis Archaeology of the Appalachian Habitation in Benton County, Highlands, edited by L. Sullivan Tennessee. Current Research in

6 The Old Man and the Pleistocene

and S. C. Prezzano, pp. 88-102. Norton, Mark R., John B. Broster, and University of Tennessee Press, Emanuel Breitburg Knoxville. 1998 The Trull Site (40PY276). Current Research in the Pleistocene 15:50- Maggard, Greg and Kacy Stackelbeck 51. 2008 Paleoindian Period. In The Archaeology of Kentucky: An Stanford, Dennis J., Elmo Leon Canales, Update - Volume One, edited by D. John B. Broster, and Mark R. Pollack, pp. 109-192. Preservation Norton Plan Report No. 3. Kentucky 2006 Clovis Blade Manufacture: Heritage Council, Frankfurt. Preliminary Data from the Carson- Conn-Short Site (40BN190), McNutt, Charles H., John B. Broster and Tennessee. Current Research in Mark R. Norton the Pleistocene 23:145-147. 2008 A Surface Collection from the Kirk Type Site (40HS174), Humphreys D. Shane Miller Mississippi State University County, Tennessee. Tennessee Archaeology 3(1):25-75. Jesse W. Tune Fort Lewis College Nami, Hugo, Mark R. Norton, Dennis J. Stanford and John B. Broster. 1996 Comments on Eastern Clovis Lithic Technology at the Carson Conn Short Site (40BN190), Tennessee River Valley. Current Research in the Pleistocene 13:62-64.

Norton, Mark R. and John B. Broster 1992 40HS200: The Nuckolls Extension Site. Tennessee Anthropologist 17:13-32. 1993 Archaeological Investigations at the Puckett Site (40SW228): A Paleoindian/Early Archaic Occupation on the Cumberland River, Stewart County, Tennessee. Tennessee Anthropologist 18:45- 58. 2008 An Overview of Paleoamerican Lithics at the Carson-Conn-Short Site (40BN190), Benton County, Tennessee. Current Research in the Pleistocene 25:125-127.

7 A RETROSPECTIVE PEEK AT THE CAREER OF JOHN BERTRAM BROSTER

Michael C. Moore, Kevin E. Smith, Aaron Deter-Wolf, and David E. Stuart

This work presents an overview of the life and archaeological career of John Bertram Broster. Few people have equaled John’s diverse experiences in archaeology, from his initial exploits on Mississippian sites in the Nashville area, through his graduate and early professional work in New Mexico (with side stints in Mexico, Europe, and Colorado), and concluding with his long and distinguished service with the Tennessee Division of Archaeology. John’s legacy to Tennessee archaeology, aside from side-splitting tales, includes his seminal research on the Paleoindian record through explorations at such sites as Carson-Conn-Short (40BN190), Coats- Hines (40WM31), and Johnson (40DV400).

The authors were privileged to be a house in Lewisburg as his dad commuted part of the 2013 Southeastern to work in Nashville. The family eventually Archaeological Conference symposium moved to Nashville, where it didn’t take honoring John Broster. Each one of us, as John long to blend in with the well as many of the folks reading this neighborhood children. For example, he article, has obtained prolific insight and showed one kid how effective a knowledge through working with John hangman’s noose can be, although the over the course of our own careers. We boy’s mother saw what was happening might add that a reasonable percentage and ran to the rescue. Yet, John’s of this insight and knowledge is actually childhood could not be considered all fun related to archaeology. A variety of and games. When John was eight years subjects arise in one’s mind when it old, he was hit by a delivery truck and comes to John, including (but not limited suffered a concussion along with other to) guns, coins, boxing, hockey, and severe injuries that put him in a body cast. Mexico. This article will attempt (but not It’s possible that accident goes a long always succeed) to focus on the way towards explaining John’s future outstanding archaeological career of John actions and behavior. Undoubtedly other Bertram Broster. events served to influence John’s outlook on life. For example, John and the senior Early Years author shared a common experience endured by select Nashville children. That John was born May 17, 1945 in would be attending Fortnightly Club, a Tallahassee, Florida to Roy and Mary charm school where 5th and 6th grade Anne Broster. Roy served as a fighter boys and girls from upstanding families pilot in Italy during World War II, and later were taught social dance and manners. enjoyed a successful career in banking. Although John and senior author Moore Mary Anne was a homemaker with a are twelve years apart in age, we passion for painting. John’s family moved discovered the instructor (Hank Fort) and to Tennessee when he was very young, teaching methods had remained the and he lived for a time in his grandfather’s same. An alleged highlight of the

8 Retrospective Peek

Fortnightly experience was the formal memories during his brief stint. Perhaps dance held every other month. All the most lasting was sustaining a serious students had to wear their Sunday-best, knee injury while playing for the USMMA including white gloves for girls. The girls football team. This particular injury could also had dance cards around their wrists be interpreted as the proverbial “glass with a line for each of the evening’s 10 half-full” as it kept him out of the Vietnam dances. Boys would have to approach the War. girls and ask for a dance, and sign the Upon John’s return to Nashville, he card if there was an opening. Usually by spent the next several years attending the fourth or fifth dance, none of the boys Middle Tennessee State College (MTSC) could remember whose card they had and Peabody Demonstration School, and signed. However, the authors are still eventually received a BA degree in trying to imagine the look on each girl’s Sociology/Anthropology from Vanderbilt face when John Broster approached to University in 1968. It was during these sign their dance card. years that John began working on John attended Hillsboro High School in archaeological sites in the Nashville area Nashville, although he actually lived in concert with Vanderbilt and the outside that school zone. Hillsboro offered Southeastern Indian Antiquities Survey Spanish classes whereas his zone school (SIAS), an organization founded by local did not, so John was able to attend this amateur archaeologists that included Bob more prestigious institution. His activities Ferguson, Buddy Brehm, and Roy Broster at Hillsboro were no different than any (Dowd and Smith 2008). This organization other student, as he lingered in the school is better known today as the Middle parking lot, played football, and visited the Cumberland Archaeological Society school library. Well, perhaps a little (MCAS). John worked on several different, but decorum prevents further significant Mississippian period sites in elucidation of his high school career. Just the Nashville area (Figure 1) including suffice it to say that John managed to Ganier, Arnold, and Sandbar Village graduate on schedule. (Broster 1972, 1988; Dowd and Broster 2012; Ferguson 1972; Smith and Moore Post High-School Activities 2013). During this period of time, John also Following his high school graduation in worked several seasons in southern 1963, John received an appointment to Mexico on archaeological projects the United States Merchant Marine directed by Ronald Spores of Vanderbilt Academy (USMMA) at Kings Point, New (Figure 2). The summers of 1966 and York. His Merchant Marine experience 1967, and winter/spring of 1970, were was short-lived, however, as John and the spent in the Nochixtlan Valley of Oaxaca. academy commandant agreed (for a John recounts many fine memories of his number of reasons) that one semester time in Mexico, although modesty was plenty of time to discern that John prevents any of these memories to be should follow a different career path. repeated in this work, with the exception While John boasts of setting the all-time of John contracting hepatitis and returning record for demerits, that particular statistic to the United States weighing about 130 is not available for review on the USMMA lbs. website. He did manage to make a few

9 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 1. John Broster in 1969 at what would later be known as 40DV36, Sandbar Village.

FIGURE 2. John (second from left) in Mexico with Ronald Spores (far right).

10 Retrospective Peek

Graduate School

Following his graduation from Vanderbilt, John headed west to attend graduate school at the University of New Mexico. It was there that John met two fellow grad students destined to become lifelong friends: most notably co-author and co-conspirator Dave Stuart who most recently served as interim director of the School for Advanced Research in Santa Fe (Stuart 2005), and some dude named Dennis Stanford. John took a number of classes from the infamous Lew Binford but never worked under him in the field. John did, however, help Binford build an adobe house for Binford’s father-in-law. In 1971, upon receiving his MA in Anthropology, John traveled to Europe and spent four months on the crew of a University of Michigan excavation in the Netherlands. By 1973, John had completed all coursework toward his PhD, but was looking for other opportunities. He spent the spring and summer of 1973 working with Stanford on the Smithsonian- FIGURE 3. John (at left) during 1973 work at sponsored excavation of the Jones-Miller the Jones-Miller site, Colorado. site in northeastern Colorado (Stanford have undoubtedly been a fatal accident. 1974, 1978), as well as excavating the Ranch 6 site in northeast New Mexico Back to Tennessee (Figure 3). The Jones-Miller excavation can’t be mentioned without relating an John had arranged to return to event that should have killed a number of Tennessee by September 1973 to begin the field crew. The crew had dug a very work with the Tennessee Division of deep trench outside the bone bed to Archaeology. The Division was conduct a geomorphological analysis. legislatively established in 1970 under the Everyone had just gotten out of the trench Tennessee Department of Conservation, and was making their way to lunch when but did not begin hiring staff a loud rumble got their attention. It turns archaeologists until 1972 (Conservation out that one entire wall of the trench had merged with the Department of collapsed, twisting the ladder left in the Environment in 1991 to create the current trench like a pretzel. While perhaps just Department of Environment and another day in the life of John Broster, Conservation). Division director Mack sheer dumb luck had averted what would Prichard hired three regional

11 Tennessee Archaeology 8(1-2) Summer 2016

also conducted investigations at the Pierce site, a Paleoindian occupation just east of Pinson in Chester County (Broster 1982). In addition to his regular Division duties, John was able to practice his pistol marksmanship on occasion, whether targeting prowlers up to no good or rats on his bedroom dresser.

Back to New Mexico

John’s service with the Division of Archaeology ended in December 1975 when he returned to the University of New Mexico to work with the Office of Contract Archaeology. While a 200-mile pipeline survey consumed much of John’s time, it was the 1976 Cochiti Lake project in northern New Mexico where John encountered perhaps his closest brush with death. During Thanksgiving week, John and a female co-worker had stayed to man the project field camp when an unexpected snowstorm dumped several feet of snow and dropped the FIGURE 4. Three original Tennessee Division of temperature that had been around 60 Archaeology regional archaeologists, circa 1974. degrees down to 25 degrees below From left to right, Carl Kuttruff, John Broster, and zero. They were trapped in the camp Brian Butler. for several days, and did everything archaeologists in 1973, with one to serve they could to survive. Well, maybe not each grand division of the state. John was everything they could. When John stated hired as the West Tennessee regional they should remove their clothes and archaeologist, with Carl Kuttruff and Brian huddle in a sleeping bag for warmth (a Butler brought on board as Middle and perfectly reasonable suggestion given the East Tennessee regional archaeologists, dire circumstances), the co-worker replied respectively (Figure 4). something to the effect that she would John was eventually stationed at rather be found frozen to death than get in Pinson Mounds State Park, living in a that sleeping bag with John. Fortunately, small farmhouse on park property (Figure brave colleagues risked their own lives in 5). He performed a number of surveys treacherous conditions to rescue John across west Tennessee, and directed the and the co-worker. Both suffered from 1974 and 1975 Memphis State University hypothermia and frostbite but eventually (now the University of Memphis) recovered. archaeological field schools at Pinson. He

12 Retrospective Peek

FIGURE 5. John and best buddies at Pinson Mounds residence, 1974.

FIGURE 6. BIA survey on Mescalero Apache Reservation, 1979.

13 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 7. 1980 School for American Research seminar, Santa Fe, New Mexico. Top row (left to right): Ed Ladd, Bruce Harrill, Fred Plog, Linda Cordell, Dee Green, Evan DeBloois. Bottom row (left to right): John Broster, Dave Stuart, Joseph Winter, James Mueller, Richard Hevely. In late 1977, John joined the Bureau of promoted to director of the Forest Indian Affairs Forest Archaeology Archaeology program in 1981, and served program. His work on tribal reservations in that capacity until the program was during this time included three surveys discontinued in 1983 under President above 8000 ft. elevation (Figure 6). This Reagan’s administration. particular research was the focus of an Soon afterward in 1984, John October 1980 School for American partnered with several individuals to form Research (SAR, later renamed the School a private consulting firm called San Juan for Advanced Research) seminar titled Basin Archaeological Consultants. The High Altitude Adaptations in the firm conducted several pipeline surveys Southwest. Participants in this seminar, in over the next year or so, but in 1985 John addition to John and co-author Dave left the company to again return to Stuart, were (in alphabetical order) Linda Tennessee. Cordell, Evan DeBloois, Dee Green, Bruce Harrill, Richard Hevely, Ed Ladd, Back (Again) to Tennessee James Mueller, Fred Plog, and Joseph Winter (Figure 7). The results of this John and his wife Diane moved back seminar were later published by the to Tennessee, and John rejoined the Forest Service (Winter 1983). John was Tennessee Division of Archaeology staff

14 Retrospective Peek

FIGURE 8. 1985 Fernvale excavation, from left to right: Bob Jolley, John Broster, and Carl Kuttruff.

FIGURE 9. John and Mike Moore during 40DV35 site visit, fall 1986.

as a crew member of the Fernvale In October 1986, John traveled to excavation project (Deter-Wolf 2013) co- England to present a paper titled Paleo- directed by Carl Kuttruff and Bob Jolley Indian Use of High Altitudes in the (Figure 8). By January 1986, John had American Southwest: Unique Training for been promoted to SHPO Archaeologist, a Cultural Resource Management position he held until June 1986 when he Archaeologists at the 11th World became the Middle Tennessee Regional Archaeological Congress in Southampton Archaeologist. It was at this time the and London (Broster 1986). This senior author was hired as SHPO presentation marked one of the last Archaeologist and received his first southwest-related activities of John’s introduction to John (Figure 9). career.

15 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 10. Mark Norton and John during early years of the Tennessee Paleoindian Projectile Point and Site Survey.

During the late 1980s, John and fellow Americas (PIDBA) maintained at the Division archaeologist Mark Norton University of Tennessee in Knoxville (Figure 10) launched their seminal project (Anderson 2013). Former Division known as the Tennessee Paleoindian archaeologists Emanuel Breitburg and the Projectile Point and Site Survey (Breitburg late Gary Barker deserve special and Broster 1994, 1995; Broster 1989; recognition for their collaborative efforts Broster and Norton 1990a, 1996; Broster on select site investigations during the et al. 1991a, 2013). To date, this project course of this project. has successfully recorded and reported The grand success of the paleo survey on many of the significant Paleoindian can be summed up in one word, trust. sites known in the state, including John and Mark have managed to develop Johnson, Coats-Hines Mastodon, Carson- lasting relationships with the amateur and Conn-Short, Widemeier, Trull, collector communities through an honest Twelkemeier, Nuckolls Extension, approach as well as lots of hard work at Sinclair, and Burgess-Mabry (Barker and society presentations and artifact shows Broster 1996; Breitburg et al. 1996; (Figure 11). These relationships have Broster and Barker 1992; Broster and provided countless leads regarding Norton 1990b, 1993, 1999, 2009; Broster previously unknown Paleoindian sites and et al. 1991b, 1994, 2006, 2008; Norton artifacts. A fact that should not go and Broster 1992, 2006, 2008; Norton et unnoticed is many of these leads had al. 1998, 2011). The results of this project been closely guarded secrets until shared have been disseminated through with John. In addition, John insisted on numerous journal articles, book chapters, sharing the credit for discoveries and and conference presentations. Also, in the excavation results with the amateur roughly 30 years since this project first archaeologists and collectors. Whether started, nearly 5500 projectile points coming up with site names or publication across the state have been documented, authorships, John made sure to with the information made available acknowledge the folks who trusted him through the Paleoindian Database of the enough to part with their information or

16 Retrospective Peek

FIGURE 11. John with Aaron Clement.

FIGURE 12. John assisting with documentation of Mississippian structure floor exposed in road cut trench at 40DV620 (Ganier Tract site), fall 2012.

provide valuable assistance. These folks assisted with Division investigations at include (in alphabetical order) Richard Gordontown, Rutherford-Kizer, Sogom, Anderson, Dennis Burgess, Kit Carson, Moss-Wright, Brentwood Library, and Aaron Clement, Gary Conn, Bobby Hulan, most recently the Mississippian David Johnson, Larry Mabrey, Rex component at the Ganier Tract site Moore, Hal Short, and Ross Sinclair. (Moore 2005; Moore and Smith 2001; John’s archaeological work in Moore et al. 2006; Norton and Broster Tennessee has not been limited to the 2004). Perhaps adding insult to injury is Paleoindian period. He’s not particularly that John’s work with the senior author in motivated to highlight the fact that he has October 2012 to document the exposed as much experience on Mississippian Mississippian features at Ganier Tract sites as anyone around (and more than comprised one of John’s last field projects most). In addition to his early work at (Figure 12). Ganier, Arnold, and Sandbar Village, John

17 Tennessee Archaeology 8(1-2) Summer 2016

Despite John’s best attempts to make us believe otherwise, he was always willing to share his knowledge and interpretations with students, professional and avocational archaeologists, and anyone else with an interest in prehistory. One special example is that of a young Vanderbilt graduate student named Kevin Smith that was interested in Middle Cumberland Mississippian sites. Ron Spores turned Kevin over to John (as well as Carl Kuttruff and John Dowd) who agreed to serve as a reader on Kevin’s dissertation (Smith 1992). While possibly a senior moment on John’s part, his assistance and guidance to other students in the two decades to follow clearly proved that beneath the gruff exterior of John Broster lay an individual constantly seeking opportunities to improve Tennessee’s archaeological record. The Tennessee Council for Professional Archaeology (TCPA) presented John with the 2013 Career Achievement Award for Professional Archaeology to celebrate his many Figure 13. TCPA President Kevin Smith successes and triumphs (Figure 13). presents John with the 2013 TCPA Career TCPA President Kevin Smith bestowed Achievement Award. the award with these opening remarks: The authors would be remiss to not mention a passion of John’s that, for a It is…my pleasure to present a Career moment anyway, was tangentially related Achievement Award for Professional to archaeology. The National Hockey Archaeology in Tennessee to someone League (NHL) Nashville Predators began who has earned our respect through nearly as an expansion franchise in 1997, and five decades of service to our profession. the mascot name and logo derive from the We will find few publications on 1971 discovery of a saber-tooth tiger in Paleoindians in Tennessee that do not cite the work of this individual. Indeed, we will downtown Nashville (Dowd 2010). This find few major publications on discovery was investigated by Vanderbilt Paleoindians in general that do not University and the SIAS. To make a long acknowledge his contributions of the past story short, the Division of Archaeology several decades…Over the past 25 years staff participated in a 1997 trailer to or so, I learned a great deal from John promote the sale of season tickets for the about field archaeology, what a Clovis new team. In this trailer, an area near the point looked like, and I may have picked up Coats-Hines Mastodon site was the scene a few choice words from him as well. of an archaeological investigation that ultimately discovers a pair of ice skates

18 Retrospective Peek

and a puck with the Predators logo. A professional actor was the primary focus of the trailer, with the rest of us filling out the background. John did have a brief speaking part, but was partially overdubbed on the final product. Over the next few years, John served as an off-ice official for the NHL coordinating media commercials with time outs at Predator games. The senior author recalls one game where John was holding a towel to the back of his head. Turns out he had slipped on the ice and opened up a nice gash. But, the team doctor sewed him up and he returned to his job duties--just another day in the life of John Broster.

Closing Statements

The authors have admittedly had a little fun at John’s expense during this work. And he certainly deserved some payback for the roles he played in our own lives. But, this amusement has in no way diminished our collective respect for John’s many accomplishments over an FIGURE 14. A smiling John Broster knows his archaeological career spanning nearly 50 retirement is just a few months away. years. For the record, however, we feel and to this day remain unexplained. We his most important accomplishment was were saddened to lose him in Oaxaca, but actually living long enough to enjoy our loss certainly was to the advantage of retirement (Figure 14). paleoindian and archaic prehistory in To close, we offer the elegant words of Tennessee and North America. John Ron Spores written for John’s retirement taught us a lot, especially during those late from the Division of Archaeology: night conferences liberally lubricated with medium-grade mescal…All praise unto the Many fond memories of John and his name and achievements of John B. wonderful parents…little may be known of Broster. his notable achievements during our surveys and excavations in the Nochixtlan Acknowledgements. The authors thank John Valley of Oaxaca, Mexico in the mid- Dowd, Mark Norton, Mack Prichard, Ron Spores, 1960s—perhaps John is trying to forget. Middle Cumberland Archaeological Society, Paleoindian Database of the Americas, and the Not only did he contribute significantly to a Tennessee Council for Professional Archaeology totally new understanding of 2500 years of for their contributions toward this work. Mixtec cultural development, but set new standards for consumption of soft drinks in one day. His mysterious trips… to Oaxaca on the weekends were a major component of our project in Yucuita and Yanhuitlan,

19 Tennessee Archaeology 8(1-2) Summer 2016

References 1986 Paleo-Indian Use of High Altitudes in the American Southwest: Unique Anderson, David G. Training for Cultural Resource 2013 PIDBA, The Paleoindian Database Management Archaeologists. of the Americas, updated 2013. Paper presented to the 11th World http://pidba.utk.edu/main.htm. Archaeological Congress in Accessed 19 May 2016. Southampton and London, England. Barker, Gary and John B. Broster 1988 Burial Patterns for the 1996 The Johnson Site (40DV400): A Mississippian Period in Middle Dated Paleoindian and Early Tennessee. Tennessee Archaic Occupation in Tennessee’s Anthropologist 13(1):1-15. Central Basin. Journal of Alabama 1989 A Preliminary Survey of Archaeology 42(2):97-153. Paleoindian Sites in Tennessee. Current Research in the Breitburg, Emanuel and John Broster Pleistocene 6:29-31. 1994 Paleoindian Site, Lithic, and Mastodon Distributions in Broster, John B. and Gary L. Barker Tennessee. Current Research in 1992 Second Report of Investigations at the Pleistocene 11:9-11. the Johnson Site, 40DV400: The 1995 The Clovis and Cumberland 1991 Field Season. Tennessee Projectile Points of Tennessee: Anthropologist 17(2):120-130. Quantitative and Qualitative Attributes and Morphometric Broster, John B. and Mark R. Norton Affinities. Current Research in the 1990a Paleoindian Fluted Point and Site Pleistocene 12:4-6. Survey in Tennessee: the 1989 Season. Current Research in the Breitburg, Emanuel, John B. Broster, Pleistocene 7:5-7. Arthur L. Reesman, and Richard G. 1990b Lithic Analysis and Paleo-Indian Stearns Utilization of the Twelkemeier Site 1996 The Coats-Hines Site: Tennessee’s (40HS173). Tennessee First Paleoindian-Mastodon Anthropologist 15:115-131. Association. Current Research in 1993 The Carson-Conn-Short Site the Pleistocene 13:6-8. (40BN190: An Extensive Clovis Habitation in Benton County, Broster, John B. Tennessee. Current Research in 1972 The Ganier Site: A Late the Pleistocene 10:3-5. Mississippian Village on the 1996 Recent Paleoindian Research in Cumberland River. In, The Middle Tennessee. In The Paleoindian Cumberland Culture, edited by and Early Archaic Southeast, Robert B. Ferguson, pp. 2-49. edited by David G. Anderson and Vanderbilt University Publications Kenneth E. Sassaman, pp. 263- in Anthropology No. 3, Nashville. 288-297. University of Alabama 1982 Paleo-Indian Habitation at the Press, Tuscaloosa. Pierce Site (40CS24); Chester 1999 A Note on OCR Dates from the County, Tennessee. Tennessee Carson-Conn-Short Site Anthropologist 7(2):93-104.

20 Retrospective Peek

(40BN190), Benton County, the Pleistocene 25:64-65. Tennessee. Current Research in the Pleistocene 16:15-16. Broster, John B., Mark R. Norton, D. 2009 The Sinclair Site (40WY111): A Shane Miller, Jesse W. Tune, and Clovis Quarry along the Buffalo Jon D. Baker. River in Wayne County, 2013 Tennessee’s Paleoindian Record: Tennessee. Current Research in The Cumberland and Lower the Pleistocene 26:35-36. Tennessee River Watersheds. In In the Eastern Fluted Point Tradition, Broster, John B., Mark R. Norton, and edited by Joseph A.M. Gingerich, Richard Anderson pp. 299-314, The University of 1991a Clovis and Cumberland Sites in the Utah Press, Salt Lake City. Kentucky Lake Region. Current Research in the Pleistocene 8:10- Deter-Wolf, Aaron 12. 2013 Fernvale (40WM51): A Late Archaic Occupation Along the Broster, John B., David P. Johnson, and South Harpeth River in Williamson Mark R. Norton County, Tennessee. Research 1991b The Johnson Site: A Dated Clovis- Series No. 19. Tennessee Cumberland Occupation in Department of Environment and Tennessee. Current Research in Conservation, Division of the Pleistocene 8:9-10. Archaeology, Nashville.

Broster, John B., Mark R. Norton, Dennis Dowd, John T. J. Stanford, C. Vance Haynes, Jr., 2010 The Nashville Smilodon: An and Margaret A. Jodry Account of the 1971 First American 1994 Eastern Clovis Adaptations in the Center Site Investigations in Tennessee River Valley. Current Davidson County, Tennessee. Research in the Pleistocene 11:12- Tennessee Archaeology 5(1):65- 14. 82.

Broster, John B., Mark R. Norton, Bobby Dowd, John T. and John B. Broster Hulan, and Ellis Durham 2012 Cockrills Bend Site 17C: A Reprint 2006 A Preliminary Analysis of Clovis from the SIAS Journal 1972. through Early Archaic Components Tennessee Archaeology 6(1-2):94- at the Widemeier Site (40DV9), 103. Davidson County, Tennessee. Tennessee Archaeology 2(2):120- Dowd, John T. and Kevin E. Smith 127. 2008 The Southeastern Indian Antiquities Survey (1963–1975) Broster, John B., Mark R. Norton, Bobby and the Middle Cumberland Hulan, and Ellis Durham Archaeological Society (1976-). 2008 Paleoindian and Early-Archaic Middle Cumberland Archaeological Occupations of the Widemeier Site Society, Nashville. (40DV9), Davidson County, Tennessee. Current Research in

21 Tennessee Archaeology 8(1-2) Summer 2016

Ferguson, Robert B. Preliminary Data from the Carson- 1972 Arnold Village Site Excavations of Conn-Short Site (40BN190), 1965-1966. In The Middle Tennessee. Current Research in Cumberland Culture, edited by the Pleistocene 23:145-147. Robert B. Ferguson, pp. 2-49. 2008 An Overview of Paleoindian Lithics Vanderbilt University Publications at the Carson-Conn-Short Site in Anthropology No. 3, Nashville. (40BN190), Benton County, Tennessee. Current Research in Moore, Michael C. the Pleistocene 25:125-127. 2005 The Brentwood Library Site: A Mississippian Town on the Little Norton, Mark R., John B. Broster, and Harpeth River, Williamson County, Emanuel Breitburg Tennessee. Research Series 15. 1998 The Trull Site (40PY276): A Tennessee Department of Paleoindian-Mastodon Association Environment and Conservation, in Tennessee. Current Research in Division of Archaeology, Nashville. the Pleistocene 15:50-51.

Moore, Michael C., Emanuel Breitburg, Norton, Mark R., John B. Broster, Dennis Kevin E. Smith, and Mary Beth Burgess, and Larry Mabrey Trubitt 2011 The Burgess-Mabrey Site: 2006 One Hundred Years of 40JK267, Jackson County, Archaeology at Gordontown: A Tennessee. Current Research in Fortified Mississippian Town in the Pleistocene 28:67-68. Middle Tennessee. Southeastern Archaeology 25(1):89-109. Smith, Kevin E. 1992 The Middle Cumberland Region: Moore, Michael C. and Kevin E. Smith Mississippian Archaeology in North 2001 Archaeological Excavations at the Central Tennessee. Ph.D. Rutherford-Kizer Site: A dissertation, Vanderbilt University, Mississippian Mound Center in Nashville. Sumner County, Tennessee. Research Series No. 13, Smith, Kevin E. and Michael C. Moore Tennessee Department of 2013 Changing Interpretations of Environment and Conservation, Sandbar Village (40DV36): Division of Archaeology, Nashville. Mississippian Hamlet, Village or Mound Center? Tennessee Norton, Mark R. and John B. Broster Archaeology 6(1-2):104-137. 1992 40HS200: The Nuckolls Extension Site. Tennessee Anthropologist Stanford, Dennis J. 17(1):13-32. 1974 Preliminary Report on the 2004 The Sogom Site (40DV68): A Excavation of the Jones-Miller Hell Mississippian Farmstead on Gap Site, Yuma, Colorado. Cockrill Bend, Davidson County, Southwestern Lore 40(3-4):29-36. Tennessee. Tennessee 1978 The Jones-Miller Site: An Example Archaeology 1(1):2-17. of Hell Gap Bison Procurement 2006 Clovis Blade Manufacture: Strategy. In, Bison Procurement

22 Retrospective Peek

and Utilization: A Symposium, edited by L. Davis and M. Wilson, pp. 90-97. Plains Anthropologist Memoir 14.

Stuart, David E. 2005 Zone of Tolerance. University of New Mexico Press, Albuquerque.

Winter, Joseph C. (editor) 1983 Cultural Resource Management: High Altitude Adaptations in the Southwest. United States Department of Agriculture, Forest Service, Southwestern Region, Report No. 2.

Michael C. Moore Tennessee Division of Archaeology 1216 Foster Ave., Cole Building #3 Nashville Tennessee 37243

Kevin E. Smith Department of Sociology and Anthropology Middle Tennessee State University Murfreesboro, Tennessee 37132-0001

Aaron Deter-Wolf Tennessee Division of Archaeology 1216 Foster Ave., Cole Building #3 Nashville Tennessee 37243

David E. Stuart Department of Anthropology University of New Mexico Albuquerque NM 87131 (retired)

23 THE PALEOINDIAN AND EARLY ARCHAIC RECORD IN TENNESSEE: A REVIEW OF THE TENNESSEE FLUTED POINT SURVEY

Jesse W. Tune

Tennessee possesses some of the densest concentrations of Paleoindian and Early Archaic artifacts in North America. As a result, it is an ideal location for research related to the early human colonization of the continent. Under John Broster’s guidance, as of 2013, there are nearly 5,500 points documented in the Tennessee Fluted Point Survey. Early Paleoindian points are the most prevalent point type recorded in Tennessee. Clovis/Gainey and Cumberland/Barnes make up over 40 percent of all Paleoindian and Early Archaic points documented in the state. The highest density of Paleoindian and Early Archaic points is recorded from the Highland Rim, and accounts for approximately two-thirds of all points in the state.

Archaeologists working in Tennessee archaeologists and the avocational have long known that the state holds a community. This paper reviews of the particularly important place in Paleoindian current status of the TFPS, and provides and Early Archaic research (e.g., Lewis a summary of statewide distributions of and Kneberg 1958; Morse et al. 1964). Paleoindian and Early Archaic points. However, nearly three decades ago John Broster lamented that, “The lack of ‘in situ’ Background sites has done much to discourage Paleoindian studies in Tennessee. The Paleoindian and Early Archaic professional community has not shown research has a long history in Tennessee. much interest in this important period” In 1945, Thomas M. N. Lewis investigated (Broster 1989:30). Unlike many western fluted points in Tennessee that he thought states, there is still a severe lack of intact to be Folsom-like. Working together with and datable early sites in Tennessee. Madeline Kneberg, they continued to do While there are exceptional numbers of Paleoindian research for several decades. Paleoindian and Early Archaic artifacts In an attempt to locate intact fluted point recorded across the state, nearly all have deposits, Lewis and Kneberg (1958) been recovered from disturbed or surface investigated the Nuckolls site, along the contexts. As a result, archaeologists Lower Tennessee River, and documented outside of the state have largely an extensive surface collection of overlooked the research potential here. Paleoindian and Early Archaic materials. Due to Broster’s dedication to In 1964 Morse and colleagues published Tennessee archaeology, and research the first map of the statewide fluted point into the Paleoindian and Early Archaic distribution in Tennessee, which included record, Tennessee is now becoming 278 points. By 1983, Guthe reported that widely recognized as a key area for 389 fluted points had been recovered in research related to the early colonization Tennessee. Like Morse and colleagues, of North America. Furthermore, under Guthe also made the prediction that this Broster’s guidance, the success of the number represented only a small portion Tennessee Fluted Point Survey (TFPS) of the fluted points that would eventually stands as an example of the importance be documented from Tennessee. of collaboration between professional In 1988 John Broster, with the

24 Tennessee Fluted Point Survey

Tennessee Division of Archaeology technological and morphological-based (TDOA), began intensively expanding the chronologies (Anderson and Sassaman TFPS, with the goal of documenting 1996; Anderson et al. 2010, 2015; Bradley Paleoindian and Early Archaic sites and et al. 2008; Ellis and Deller 1997; collections throughout the state (Broster Goodyear 1999; Meltzer 2009; Tankersley 1989). In turn, Broster used the data 1990, 1996; Tune 2016; see Gramly 2013 compiled in the survey to help construct for alternative). regional settlement models to further Extensive fluted point survey data assess mobility and land use strategies have been collected by statewide surveys during the late Pleistocene (Broster and and compiled in PIDBA (Anderson 2004; Norton 1996). The first report of the Anderson et al. 2010; Goodyear 1999; TFPS, a year into its existence, included Miller and Gingerich 2013). Potential 105 locations where fluted points had biases and limitations are well known for been recovered (Broster 1989). This PIDBA datasets such as incomplete data, included 58 sites where multiple sample inconsistency, site formation Paleoindian artifacts were recovered, and processes, and ground cover (Anderson 47 locations of isolated Paleoindian et al. 2010; Ballenger et al. 2011; artifacts. Prasciunas 2011). However, such data Under Broster’s guidance, the TFPS are commonly accepted for modeling has become one of the most certain human behaviors (Anderson and comprehensive statewide surveys in the Gillam 2000; Anderson et al. 2011; Lanata country. Thanks to his dedication and et al. 2008; Meeks and Anderson 2012; perseverance, the TFPS has been greatly Miller 2011; Shott 2013; Smallwood 2012; expanded and now includes nearly 5,500 Smallwood et al. 2015; Tune 2016). Paleoindian and Early Archaic points. This survey is exceptional because, not only Methods does it differentiate between Paleoindian point types, but it also includes Early The distribution of Paleoindian and Archaic point forms. Adding to the Early Archaic points presented here is significance of the TFPS is the fact that is based on the 2013 TFPS update. Spatial has been made accessible for research comparisons were made based on through the Paleoindian Database of the county-level provenience and assessed Americas (PIDBA). by frequencies and densities. Physiographic comparisons were made Fluted Point Survey Data by grouping counties into eight major physiographic regions: Alluvial Plain, Stratified, radiocarbon-datable Paleo- Coastal Plain, Highland Rim, Central indian and Early Archaic sites are Basin, Cumberland Plateau, Cumberland notoriously rare in Tennessee and the Mountains, Ridge and Valley, and Blue greater southeastern United States Ridge Mountains (Figure 1; Fenneman (Anderson 2005; Anderson et al. 2015; 1917). As county boundaries do not Dunnell 1990; Goodyear 1999; Miller and always align with physiographic Gingerich 2013). As such, absolute boundaries, some counties fall within chronologies are not addressed here. multiple physiographic regions. To Rather, the relative typological sequence address this situation, the area (km2) of used here follows the widely accepted each county was calculated for each

25 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 1. Map of the state of Tennessee showing the counties and physiographic regions referenced in this paper. physiographic region using ArcGIS 10.2. by John Broster and Mark Norton. Due to In turn, the percentage of each county typological uncertainty that currently was used to estimate the number of surrounds the distinction between Clovis points within each physiographic region. and Gainey (e.g., Eren et al. 2011; The overall densities for each point type Morrow and Morrow 2002), and are recorded by physiographic region. To Cumberland and Barnes (e.g., Bradley et account for the differing sizes of each al. 2010; Justice 1987; White 2006), the physiographic region, point densities were typology used here groups these forms scaled to 1,000 km2. Scaling the densities into two types. Clovis and Gainey points provides a more accurate reflection of are combined into a single type, as are statewide point distributions, and enables Cumberland and Barnes points. comparisons between physiographic regions. Paleoindian and Early Archaic Point To address the statewide distributions Distributions of Paleoindian and Early Archaic points, artifacts are grouped into eight previously As of 2013, a total of 5,497 defined types. Paleoindian types consist Paleoindian and Early Archaic points are of Clovis/Gainey, Cumberland/Barnes, recorded in the TFPS, including Redstone, Quad, Beaver Lake, and unspecified fragments and preforms. Of Dalton, while Early Archaic types include that total number, only points clearly Harpeth River and Greenbrier (Broster et identified to a single type with county-level al. 2013; Cambron 1970; Justice 1987). It provenience are used in this study, which should be noted that some debate exists leaves 2,712 points in the revised dataset. as to the interpretation of Dalton. As this This includes 687 Clovis/Gainey, 451 type has been previously dated to span Cumberland/Barnes, 77 Redstone, 289 the Pleistocene/Holocene boundary, it has Quad, 340 Beaver Lake, 316 Dalton, 103 been referred to as Transitional Harpeth River, and 449 Greenbrier (Table Paleoindian by some researchers 1). (Clayton 1965; DeJarnette et al 1962; Plotting the distribution of points by Driskell et al. 2012; Goodyear 1982; county reveals clear patterns in statewide Sherwood et al. 2004). densities. While Paleoindian point types All typological identifications of points are generally found throughout the entire included here were originally documented state, Early Archaic points appear more

26 Tennessee Fluted Point Survey restricted in distribution. The Cumberland sections of rivers, ecotones, and lithic Mountain region is the only region where sources, and may reflect land-use Paleoindian and Early Archaic points are strategies (Miller 2011). As such, the not currently documented. As such, this Lower Tennessee River may have simply region is omitted from further analyses. been an ideal place to live during the Two counties have markedly denser Pleistocene-to-Holocene transition. concentrations than the rest of the state, The inclusion of data from Benton and and warrant further discussion. Benton Humphreys Counties in this study would and Humphreys counties border the disproportionally privilege the Coastal confluence of the Lower Tennessee and Plain and Highland Rim regions. Table 1 Duck Rivers. This portion of the presents the statewide point frequencies Tennessee River Valley has been with and without data from Benton and significantly impacted by the Humphreys Counties. Because of the impoundment of the Kentucky Lake. As a disproportional numbers of points from result, the majority of points recovered these two counties (approximately 47% of from this area are from highly deflated the entire TFPS), they are excluded from shorelines (Broster et al. 1991, 1994; further analyses here. The exclusion of Lewis and Kneberg 1958). Furthermore, these counties further reduces the dataset the TDOA, under the direction of Broster, analyzed here to 1,796 points, and has conducted extensive research in enables a more accurate assessment of these two counties. Most notably, their the statewide distribution of points. work at Carson-Conn-Short has led to roughly 700 Paleoindian bifaces being Alluvial Plain documented, including points and The Alluvial Plain consists of at least preforms (Broster and Norton 1993, 1996; 2 Broster et al. 1994, 1996, 2013). part of six counties, totaling 2,787 km . The exceptionally high point densities Paleoindian and Early Archaic points are also warrant consideration of potential relatively rare in the Alluvial Plain. Only sampling biases (see Anderson et al. six points are documented from this 2010; Lepper 1983, 1985; Prasciunas region, accounting for 0.3 percent (%) of 2011; Seeman and Prufer 1982; Shott all Paleoindian and Early Archaic points in 2002). However, Miller (2011) has the state. There is one Clovis/Gainey, one demonstrated that selective recovery Beaver Lake, two Daltons, and two biases may not be a significant factor in Greenbriers recorded (Table 2). county-level data in the region. Rather, higher concentrations of points are frequently documented at the inter-

Table 1. Frequency of Point Types by Physiographic Region.

27 Tennessee Archaeology 8(1-2) Summer 2016

Table 2. Distribution of Points in the Alluvial Plain.

Table 3. Distribution of Points in the Coastal Plain.

Coastal Plain Henderson, Obion, and Tipton Counties. Cumberland/Barnes points are also The Coastal Plain consists of at least frequently identified in the Coastal Plain. part of 19 counties (excluding Benton There are 35 points documented, County), totaling 23,102 km2. This region representing 8.6% of all has one of the highest frequencies of Cumberland/Barnes in the state. Paleoindian and Early Archaic points in Cumberland/Barnes points are the state. There are 249 Paleoindian and documented in McNairy, Gibson, Henry, Early Archaic points currently documented Decatur, Hardin, Weakley, Carroll, in the Coastal Plain, which accounts for Chester, Haywood, and Henderson 13.9% of all points documented (Table 3). Counties. Clovis/Gainey points are particularly Eighteen Redstone points are prevalent in this region with 99 points documented in the Coastal Plain. This currently documented. This represents represents 29.5% of all Redstones 18.2% of all Clovis/Gainey points known currently known from Tennessee. They throughout the state. The majority of are documented from Carroll, Crockett, Clovis/Gainey points from the Coastal Decatur, Gibson, Hardin, Henderson, Plain come from Henry County, followed Henry, and McNairy Counties. by Carroll and Gibson Counties. The Twenty-six Quad points are known remainder are documented from Weakley, from the Coastal Plain, representing McNairy, Chester, Crockett, Hardin, 14.1% of the total number recorded Haywood, Shelby, Decatur, Dyer, throughout the state. Quads are

28 Tennessee Fluted Point Survey

Table 4. Distribution of Points in the Highland Rim.

documented in Carroll, Decatur, Gibson, from the state. Harpeth River points have Hardin, Henry, Madison, and McNairy a very constricted distribution, occurring Counties. only in Henry and McNairy Counties. Fourteen Beaver Lake points are Twenty-nine Greenbrier points come documented in the Coastal Plain. This from the Coastal Plain, representing represents 6.5% of all Beaver Lakes in 16.2% of all Greenbriers in Tennessee. the state. Beaver Lakes are known from The majority comes from Carroll County, Carroll, Crockett, Decatur, Gibson, Henry, followed by Chester, Decatur, Gibson, Lauderdale, McNairy, and Shelby Hardin, Henry, and Lauderdale Counties. Counties. Twenty-six Dalton points are Highland Rim documented from the Coastal Plain, which represents 17.1% of all Daltons from The Highland Rim is the largest Tennessee. They are documented in physiographic region in the state, and Carroll, Decatur, Gibson, Hardin, consists of at least part of 40 counties Haywood, Henry, Lauderdale, McNairy, (excluding Benton and Humphreys Obion, Shelby, Tipton, and Weakley Counties), totaling 33,040 km2. More Counties. Paleoindian and Early Archaic points are Only two Harpeth River points are documented in the Highland Rim than all known from the Coastal Plain, which other regions in Tennessee combined. represents 3.4% of all Harpeth Rivers There are currently 1,134 points

29 Tennessee Archaeology 8(1-2) Summer 2016

documented from this region, accounting counties in the region except for Smith, for 63.1% of all points in the state (Table Marshall, Henry, Trousdale, Cannon, 4). Clay, Franklin, Giles, Grundy, Hardin, Clovis/Gainey is the most common Hickman, Lincoln, Macon, Moore, Pickett, point type from the Highland Rim (n=317). Putnam, and Sequatchie. This represents 58.4% of all A total of 97 Dalton points are known Clovis/Gainey points currently from the Highland Rim, which makes up documented in Tennessee. Clovis/Gainey 63.8% of all Daltons in Tennessee. They points are recorded for all counties in the are documented from Stewart, Robertson, Highland Rim except for Grundy, Lewis, Houston, Davidson, Henry, Dickson, Macon, Pickett, Putnam, and Sequatchie. Montgomery, DeKalb, Cheatham, Smith, A total of 258 Cumberland/Barnes Coffee, Maury, Williamson, Hardin, points are documented from the Highland Marshall, Decatur, Jackson, Lawrence, Rim. This represents 63.5% of all Overton, Perry, Sumner, and Van Buren Cumberland/Barnes currently known from Counties. the state. The majority of There are 47 Harpeth River points Cumberland/Barnes points in the from the Highland Rim, representing Highland Rim are from Smith, Sumner, 81.0% of the total number known and Lawrence Counties. The only throughout the state. The majority are counties in the region where from Davidson, and Dickson Counties. Cumberland/Barnes are not documented The remainder is documented from are Cannon, Clay, Grundy, Macon, Robertson, Cheatham, Houston, Lincoln, Pickett, and Sequatchie. Montgomery, Henry, Jackson, and Twenty-four Redstones are docu- Stewart Counties. mented from the Highland Rim, making up The vast majority (75.4%) of 39.2% of all Redstones from the state. Greenbrier points in Tennessee are from These points have been documented from the Highland Rim (n=135). They are Decatur, Henry, Giles, Franklin, Hardin, documented from Houston, Benton, Smith, Sumner, Houston, Lawrence, Robertson, Davidson, Decatur, Henry, Lewis, Montgomery, Overton, Stewart, Stewart, Hardin, Jackson, Williamson, Warren, and Williamson Counties. Montgomery, Cheatham, Dickson, A total of 111 Quad points are known Franklin, Hickman, Lincoln, Maury, Perry, from the Highland Rim, accounting for Sumner, Trousdale, and Warren 60.3% of all Quads in the state. These Counties. points have been documented from Robertson, Henry, Lawrence, Bedford, Central Basin Davidson, Sumner, Dickson, Jackson, Montgomery, Smith, Cheatham, Houston, The Central Basin consists of at least Stewart, Wayne, White, Decatur, part of 18 counties, totaling 9,740 km2. Marshall, Perry, Van Buren, Williamson, This region has the second highest Coffee, Hardin, Hickman, Maury, Overton, concentration of points in the state. There and Pickett. are 261 Paleoindian and Early Archaic There are 145 Beaver Lake points points documented from this region, documented from the Highland Rim, accounting for 14.5% of the state total accounting for 67.8% of all Beaver Lakes (Table 5). in the state. They are documented in all

30 Tennessee Fluted Point Survey

Table 5. Distribution of Points in the Central Basin.

The Central Basin is the only Smith, Davidson, Sumner, Rutherford, physiographic region in Tennessee where Jackson, Marshall, and Williamson Clovis/Gainey are not the dominant point Counties. type. There are 74 Clovis/Gainey points, Beaver Lake points are relatively which account for 13.6% of the state total. common in the Central Basin. There are These are primarily documented in 41 points, accounting for 19.2% of all Davidson and Smith Counties, followed by Beaver Lakes documented throughout the Sumner, Trousdale, Bedford, Rutherford, state. They occur most frequently in Williamson, Maury, Cannon, DeKalb, Bedford and Davidson Counties, while the Jackson, and Marshall. remainder are documented from Smith, Cumberland/Barnes points are the Rutherford, Jackson, Sumner, Wilson, most frequent type documented from the Coffee, Marshall, Maury, Trousdale, and Central Basin. There are 79 Williamson Counties. Cumberland/Barnes points, representing There are 18 Dalton points 19.5% of the total sample. These are documented from the Central Basin. This most frequently document in Smith and represents 11.8% of all Daltons Sumner Counties, followed by Bedford, documented in Tennessee. They are most Davidson, Wilson, Trousdale, Marshall, frequently found in Davidson County, Maury, Rutherford, Williamson, Coffee, followed by Smith, Rutherford, DeKalb, DeKalb, and Jackson. Marshall, Maury, and Williamson. Nine Redstone points are the least common Harpeth River points are documented in type found in the Central Basin. There are the Central Basin and all come from only four Redstones from this region, Davidson County. This accounts for which accounts for 6.6% of the state total. 15.5% of all Harpeth Rivers documented They are documented from Smith, in Tennessee. Ten Greenbrier points are Rutherford, and Sumner Counties. documented from the Central Basin, There are 26 Quad points documented which represents 6.0% of all Greenbriers from the Central Basin, accounting for from Tennessee. They are found in 14.1% of the total sample. They occur in Davidson, Jackson, Marshall, Trousdale, relatively low frequencies in Bedford, and Williamson Counties.

31 Tennessee Archaeology 8(1-2) Summer 2016

Table 6. Distribution of Points in the Cumberland Plateau.

Cumberland Plateau are documented from Hamilton, Rhea, and Overton Counties. Eight Quad points The Cumberland Plateau consists of at are known from the Cumberland Plateau, least part of 18 counties, totaling 12,120 representing 4.3% of all Quads from the km2. There are 45 Paleoindian and Early state. The majority of Quads are Archaic points documented in this region, documented from Hamilton County, which represents 2.5% of the state total followed by Fentress, Van Buren, and (Table 6). White. Six Beaver Lakes (2.8%) are Clovis/Gainey and Cumberland/ documented from the Cumberland Barnes points are documented in equal Plateau. They were recovered from frequencies from the Cumberland Hamilton, Van Buren, White, and Overton Plateau, although their distributions are Counties. There are only three Daltons slightly different. There are 12 known from the Cumberland Plateau, Clovis/Gainey points, which accounts for which accounts for 2.0% of all Daltons in 2.2% of the state total. Hamilton County Tennessee. They are documented from has the highest frequency of Hamilton and Van Buren Counties. A Clovis/Gainey, followed by Bledsoe, single Greenbrier is documented from Fentress, Franklin, Morgan, Overton, Van Hamilton County, and represents 0.5% of Buren, and White. The 12 Cumberland/ all Greenbriers in the state. There are no Barnes points from the Cumberland Harpeth Rivers currently known from this Plateau represent 3.0% of all region. Cumberland/Barnes from Tennessee. The majority of Cumberland/Barnes come Ridge and Valley from Hamilton County, followed by Van Buren, Cumberland, Overton, Putnam, The Ridge and Valley region consists and White. of at least part of 29 counties, totaling Only three Redstone points are 20,066 km2. There are 90 Paleoindian and currently documented from the Early Archaic points from this region, Cumberland Plateau, accounting for 4.9% which accounts for 5.0% of the state total of all Redstones from Tennessee. They (Table 7).

32 Tennessee Fluted Point Survey

Table 7. Distribution of Points in the Ridge and Valley.

Clovis/Gainey are the dominant point documented from this region and all are type in the Ridge and Valley region. There from Hamilton County. are 34 Clovis/Gainey points, accounting There are seven Beaver Lake points for 6.2% of the state total. Hamilton from the Ridge and Valley region, which County has the highest frequency, represents 3.2% of the state total. They followed by Knox, Washington, Greene, are documented from Hamilton, Grainger, McMinn, Meigs, Sullivan, Carter, Hawkins, Greene, and McMinn Counties. There are Jefferson, Loudon, Polk, Rhea, and five Dalton points from this region, Roane. accounting for 3.3% of all Daltons in Cumberland/Barnes is the second Tennessee. They are documented from most abundant point type in the Ridge Hamilton and Greene Counties. Only a and Valley region with 20 points, which single Greenbrier has been documented account for 4.9% of the state total. The for this region and comes from Hamilton majority is documented from Hamilton County. This represents 0.5% of all County, followed by Knox, McMinn, Greenbriers in the state. There are no Meigs, Hamblen, Rhea, and Roane. Harpeth River points currently Redstone and Quad points are documented from the Ridge and Valley documented in equal numbers from this region. region with 11 points of each type. This accounts for 18.0% of all Redstones Blue Ridge Mountains documented in Tennessee. Redstone points are known from Hamilton, Knox, The Blue Ridge Mountains consists of Rhea, Hawkins, and McMinn Counties. at least part of 13 counties, totaling 6,372 Six percent of all Quads in Tennessee are km2. Paleoindian and Early Archaic points

33 Tennessee Archaeology 8(1-2) Summer 2016

Table 8. Distribution of Points in the Blue Ridge Mountains.

occur in low frequencies in the Blue Ridge frequencies, cumulatively totaling 8.4% of Mountain region of eastern Tennessee. all points. There are 11 points accounting for only Assessing the frequencies and 0.6% of the state total (Table 8). There percentages of individual point types by are six Clovis/Gainey points (1.1%) from physiographic region provides additional Polk, Greene, Johnson, and Sevier explanation (Table 9). The frequencies of Counties. Two Cumberland/Barnes points points differ significantly by physiographic (0.5%) are documented from Polk and region (X2=135.727, df=42, p=<0.001). In Sevier Counties. There is a single the Alluvial Plain all point types occur very Redstone (1.6%) from Blount County, one near expected frequencies. A similar Quad (0.5%) from Sevier, and one Dalton pattern is evident in the Blue Ridge (0.7%) from Monroe County. There are no Mountains where Clovis/Gainey occur Beaver Lake, Harpeth River, or slightly more frequently than expected, Greenbrier points documented from this but all other point types occur very near region. the expected frequencies. Clovis/Gainey occurs more frequently Point Frequency by Physiographic than expected in the Coastal Plain, Ridge Region and Valley, and the Blue Ridge Mountains, and less frequently than The overall frequencies of points by expected in the Highland Rim. physiographic region further illustrate Clovis/Gainey occur near expected patterns in statewide distributions. The frequencies in the Central Basin and vast majority of all points occurs in the Cumberland Plateau. Cumberland/Barnes Highland Rim (63.1%). The Central Basin occur more frequently than expected only (14.5%) and Coastal Plain (13.9%) have in the Central Basin, and less frequently roughly equal frequencies of points. than expected in the Coastal Plain. They Points from these three regions make up are documented at nearly expected over 90% of all Paleoindian and Early frequencies in all other regions. Archaic points documented in the state. Redstones occur more frequently than The Ridge and Valley, Cumberland expected in the Coastal Plain and Ridge Plateau, Blue Ridge Mountains, and and Valley, and less frequently than Alluvial Plain have much smaller expected in the Highland Rim and Central

34 Tennessee Fluted Point Survey

Table 9. Frequency of Points by Physiographic Region.

Basin. They occur near the expected frequencies in the Highland Rim and frequency in the Cumberland Plateau. Cumberland Plateau. Harpeth Rivers only Quads occur slightly more frequently than occur more frequently than expected in expected in the Cumberland Plateau, and the Highland Rim, and only less frequently slightly less frequently than expected in than expected in the Coastal Plain. They the Highland Rim. They are near occur very near the expected frequencies expected frequencies in the Coastal Plain, in all other regions. Greenbriers occur at Central Basin, and Ridge and Valley. higher than expect frequencies in the Beaver Lakes occur more frequently than Highland Rim and Coastal Plain, and expected in the Highland Rim and Central lower than expected in all other regions. Basin, and less frequently than expected in the Coastal Plain and Ridge and Valley. Point Densities by Physiographic They are very near the expected Region frequency in the Cumberland Plateau. Daltons occur slightly more frequently To further interpret the distribution of than expected in the Coastal Plain, and Paleoindian and Early Archaic points, the slightly less frequently than expected in densities of points were scaled to account the Central Basin and Ridge and Valley. for differing sizes of physiographic regions They occur at nearly expected (Table 10). The cumulative density of all

35 Tennessee Archaeology 8(1-2) Summer 2016

Table 10. Density of Points (per 1,000 km2) by Physiographic Region.

Paleoindian and Early Archaic points densities. throughout the state is 84.00 points per Harpeth River points occur at 1,000 km2. The overall pattern approximately one point per 1,000 km2 in corresponds well to the statewide county the Highland Rim (1.42 per 1,000 km2) densities. The highest density of each and Central Basin (0.97 per 1,000 km2). point type occurs in the Highland Rim. Greenbrier has the highest density in the The density of Clovis/Gainey points in the Highland Rim (4.09 per 1,000 km2) Highland Rim is 9.59 points per 1,000 followed by the Coastal Plain (1.26 per km2, followed by the Central Basin (7.57 1,000 km2) and Central Basin (1.08 per per 1,000 km2), Coastal Plain (4.29 per 1,000 km2). 1,000 km2), and Ridge and Valley (1.68 per 1,000 km2). The same pattern is Conclusion evident in Cumberland/Barnes, which occur in nearly equal densities in the Thanks to a career of hard work and Central Basin (8.12 per 1,000 km2) and guidance by John Broster, an extensive Highland Rim (7.81 per 1,000 km2), dataset now exists that can be used to followed by the Coastal Plain (1.52 per address broad archaeological questions 1,000 km2) and Ridge and Valley (1.01 related to the initial human colonization of per 1,000 km2). North America. The TFPS is one of the Redstone points occur in densities of most comprehensive statewide surveys in less than one point per 1,000 km2 in all North America. Not only does this survey regions. Quad points occur in the highest differentiate between Paleoindian point density in the Highland Rim at 3.36 per types, but it also includes Early Archaic 1,000 km2, followed by the Central Basin types. While this dataset may focus on the (2.62 per 1,000 km2) and Coastal Plain archaeological record of Tennessee, it is (1.13 per 1,000 km2). All other regions relevant to research throughout North Quad occurs less than one point per America. 1,000 km2. Beaver Lake points occur in Significant differences exist in the nearly equal densities in the Highland Rim frequencies and regional densities, as (4.39 per 1,000 km2) and Central Basin well as distributions of Paleoindian and (4.17 per 1,000 km2). They occur below Early Archaic points throughout one point per 1,000 km2 in all other Tennessee. There appears to be a regions. The highest density of Dalton general reduction in the overall points occurs in the Highland Rim (2.94 frequencies of points over time in per 1,000 km2), followed by the Central Tennessee from the Paleoindian through Basin (1.82 per 1,000 km2) and Coastal the Early Archaic. However, Redstone, Plain (1.13 per 1,000 km2) at nearly equal assumed to be a Middle Paleoindian type,

36 Tennessee Fluted Point Survey

is the least common Paleoindian and suggestions, which greatly improved this paper. Early Archaic point type in Tennessee. Paleoindian and Early Archaic points References

are concentrated toward the center of the Anderson, David G. state, with an overwhelming majority of all 2004 Paleoindian Occupations in the points documented from the Highland Southeastern United States. In Rim, Coastal Plain, and Central Basin. New Perspectives on the First Over 50% of every Paleoindian and Early Americans, edited by Bradley T. Archaic point type documented in Lepper and Robson Bonnichsen, Tennessee, except Redstone, comes pp. 119-128. Texas University from the Highland Rim. While this is the Press, College Station. largest physiographic region in 2005 Pleistocene Human Occupation of Tennessee, scaling point densities to the Southeastern United States: account for differing region sizes further Research Directions for the Early supports the significance of this region. 21st Century. In Paleoamerican Determining if the apparent patterns Origins: Beyond Clovis, edited by observed in the Paleoindian and Early R. Bonnichsen, B. T. Lepper, D. Archaic record of Tennessee reflect actual Stanford, and M. R. Waters, pp. trends in prehistoric adaptations is 29–42. Center for the Study of the complicated by the lack of buried and First Americans, Texas A&M radiocarbon-datable sites. Until assumed University Press, College Station. typological chronologies are supported by

archaeological evidence, questions will Anderson, David G. and J. Gillam remain as to the relationships between 2000 Paleoindian Colonization of the individual point types. Americas: Implications form an The hard work of John Broster has Examination of Physiography, resulted in the documentation of a Demography, and Artifact substantial Paleoindian and Early Archaic Distribution. American Antiquity record in Tennessee. The success in 65:43-66. compiling such a comprehensive

statewide survey stands as an example of Anderson, David G., Albert C. Goodyear, the importance in working closely with James Kennett, and Allen West avocational groups. It is now up to others 2011 Multiple Lines of Evidence for to continue this work and build upon its Possible Human Population legacy. Decline/Settlement Reorganization

Acknowledgements. I owe a great deal of gratitude during the Early Younger Dryas. to John Broster for getting me into this mess to Quaternary International 242:570- begin with. He took the time to introduce me to 583. Paleoindian archaeology while I was at a Mississippian field school, and has helped me with Anderson, David G., D. Shane Miller, numerous projects since then. I have learned many, many things from John over the years. I Stephen J. Yerka, J. Christopher would like to thank Mike Moore and Kevin Smith Gillam, Erik N. Johanson, Derek T. for the opportunity to write this paper and Anderson, Albert C. Goodyear, and subsequently organize this special issue of Ashley M. Smallwood Tennessee Archaeology in honor of John Broster’s 2010 PIDBA (Paleoindian Database of career and influence. I also thank an anonymous reviewer for their helpful comments and the Americas) 2010: Current Status

37 Tennessee Archaeology 8(1-2) Summer 2016

and Findings. Archaeology of Indian Sites in Tennessee. Current Eastern North America 38:63-90. Research in the Pleistocene 6:29- 31. Anderson, David G. and Kenneth E. Sassaman (editors) Broster, John B. and Mark R. Norton 1996 The Paleoindian and Early Archaic 1993 The Carson-Conn-Short Site Southeast. University of Alabama (40BN190): An Extensive Clovis Press, Tuscaloosa. Habitation in Benton County, Tennessee. Current Research in Anderson, David G., Ashley Smallwood, the Pleistocene 10:3-4. and D. Shane Miller 1996 Recent Paleoindian Research in 2015 Pleistocene Human Settlement in Tennessee. In The Paleoindian the Southeastern United States: and Early Archaic Southeast, Current Evidence and Future edited by D. G. Anderson and K. E. Directions. PaleoAmerica 1(1):7- Sassaman, pp. 288-297. University 51. of Alabama Press, Tuscaloosa.

Ballenger, Jesse A. M., Vance T. Holliday, Broster, John B., Mark R. Norton, and Andrew L. Kowler, William T. Richard Anderson Reitze, Mary M. Prasciunas, D. 1991 Clovis and Cumberland Shane Miller, and Jason D. Sites in the Kentucky Lake Region. Windingstad Current Research in the 2011 Evidence for Younger Dryas Global Pleistocene 8:10-12. Climate Oscillation and Human Response in the American Broster, John B., Mark R. Norton, D. Southwest. Quaternary Shane Miller, Jesse W. Tune, and International 242:502-519. John D. Baker 2013 Tennessee’s Paleoindian Record: Bradley, Bruce A., Michael B. Collins, and The Cumberland and Lower Andrew Hemmings Tennessee River Watersheds. In In 2010 Clovis Technology. International The Eastern Fluted Point Tradition, Monographs in Prehistory, edited by Joseph A.M. Gingerich, Archaeological Series 17, Ann 299-314. University of Utah Press, Arbor. Salt Lake City.

Bradley, James W., Arthur E. Spiess, Broster, John B., Mark R. Norton, Dennis Richard A. Boisvert, and Jeff J. Stanford, C. Vance Haynes Jr., Boudreau and Margaret A. Jodry 2008 What’s the Point? Modal Forms 1994 Eastern Clovis Adaptations in the and Attributes of Paleoindian Tennessee River Valley. Current Bifaces in the New England- Research in the Pleistocene 11:12- Maritimes Region. Archaeology of 14. Eastern North America 36:119-172. 1996 Stratified Fluted Point Deposits in the Western Valley of Tennessee. Broster, John B. Proceedings of the 14th Annual 1989 A Preliminary Survey of the Paleo- Mid-South Archaeological

38 Tennessee Fluted Point Survey

Conference Special Publications 1, Eren, Metin I., Stanley Vanderlaan, and edited by Richard Walling, Camille John D. Holland Wharey and Camille Stanley, pp. 1- 2011 Overshot Flaking at the Arc Site, 11 Panamerican Consultants, Inc., Genesee County, New York: Memphis. Examining the Clovis-Gainey Connection. The Open Cambron, James W. Anthropology Journal 4:40-52. 1970 Harpeth River Point. Tennessee Archaeologist 20(1):15-18. Fenneman, Nevin M. 1917 Physiographic Subdivision of the Clayton, M. V. United States. Proceedings of the 1965 Bluff Shelter Excavations on Sand National Academy of Sciences Mountain. Journal of Alabama 3(1):17-22. Archaeology 11:1–101. Goodyear, Albert C. DeJarnette, D. L., E. B. Kurjack, and J. W. 1982 The Chronological Position of the Cambron Dalton Horizon in the Southeastern 1962 Stanfield-Worley Bluff Shelter United States. American Antiquity Excavations. Journal of Alabama 47(2):382-395. Archaeology 8:1–111. 1999 The Early Holocene Occupation of the Southeastern United States: A Driskell, Boyce N., Scott C. Meeks, and Geoarchaeological Summary. In Sarah C. Sherwood Ice Age Peoples of North America: 2012 The Transition from Paleoindian to Environments, Origins, and Archaic in the Middle Tennessee Adaptations, edited by R. Valley. In From the Pleistocene to Bonnichsen and K. L. Turnmire, pp. the Holocene, edited by C. Britt 432-481. Oregon State University Bousman and Bradley J. Vierra, Press, Corvallis. pp. 253-271. Texas A&M University Press, College Station. Gramly, Richard M. 2013 Utilized Prismatic Blades vs. Dunnell, Robert C. Utilized Flakes at the Phil Stratton 1990 The Role of the Southeast in Cumberland Site, Kentucky. North American Archaeology. American Archaeologist 34(2):137- Southeastern Archaeology 9:11– 167. 22. Guthe, Alfred K. Ellis, Christopher J. and D. Brian Deller 1983 Comments Regarding the 1997 Variability in the Archaeological Inventory of Fluted Points in Record of Northeastern Paleo- Tennessee. Archaeology of Indians: A View from Southern Eastern North America 11:23-24. Ontario. Archaeology of Eastern North America 25:1-30. Justice, Noel D. 1987 Stone Age Spear and Arrow Points of the Midcontinental and eastern United States. Indiana University

39 Tennessee Archaeology 8(1-2) Summer 2016

Press, Bloomington. Colonizing Ice Age America. University of California Press, Lanata, J., L. Martino, A. Osella, and A. Berkley. Garcia-Herbst 2008 Demographic Conditions Miller, D. Shane Necessary to Colonize New 2011 Rivers, Rocks, and Eco-Tones: Spaces: The Case for Early Human Modeling Clovis Landscape-Use in Dispersal in the Americas. World the Southeastern United States. Archaeology 40:520-537. Paper presented at the 76th Annual Meeting of the Society for Lepper, Bradley. T. American Archaeology, 1983 Fluted Point Distributional Patterns Sacramento. in the Eastern United States: A Contemporary Phenomenon. Miller, D. Shane and Joseph A. M. Midcontinental Journal of Gingerich Archaeology 8:269-285. 2013 Regional Variation in the Terminal 1985 The Effects of Cultivation and Pleistocene and Early Holocene Collecting on Ohio Fluted Point Radiocarbon Record of Eastern Finds: A Reply to Seeman and North America. Quaternary Prufer. Midcontinental Journal of Research 79:175-188. Archaeology 10:241-250. Morrow, Juliet E. and Toby A. Morrow Lewis, Thomas M. N. 2002 Exploring the Clovis-Gainey- 1945 On Folsom Arrowpoints. Folsom Continuum: Technological Tennessee Archaeologist 1(3):14- and Morphological Variation in 15. Midwestern Fluted Points. In Folsom Technology and Lifeways, Lewis, Thomas M. N. and Madeline edited by John E. Clark and Kneberg Michael B. Collins, pp. 141-157. 1958 The Nuckolls Site. Special Publication No. 4, Lithic Tennessee Archaeologist 14(2):60- Technology, Tulsa. 79. Morse, Dan F., Phyllis A Morse, and John Meeks, Scott and David G. Anderson Wagner Jr. 2012 Evaluating the Effect of the 1964 Fluted Points from Smith County, Younger Dryas on Human Tennessee. Tennessee Population Histories in the Archaeologist 10(1):16-34. Southeastern United States. In Hunter-Gatherer Behavior: Human Prasciunas, Mary M. Responses during the Younger 2011 Mapping Clovis: Projectile Points, Dryas, edited by Metin I. Eren, pp. Behavior, and Bias. American 111-138. Left Coast Press, Walnut Antiquity 76(1):107-126. Creek. Seeman, Mark F. and Olaf H. Prufer Meltzer, David J. 1982 An Updated Distribution of Ohio 2009 First Peoples in a New World: Fluted Points. Midcontinental

40 Tennessee Fluted Point Survey

Journal of Archaeology 7(2):155- Supplement 5. JAI Press, 169. Greenwich, Connecticut. 1996 Ice Age Hunters and Gatherers. In Sherwood, Sarah C., Boyce N. Driskell, The Prehistory of Kentucky, edited Asa R. Randall, and Scott C. by R. B. Lewis. University of Meeks Kentucky Press, Lexington. 2012 Chronology and Stratigraphy at Dust Cave, Alabama. American Tune, Jesse W. Antiquity 69(3):533-554. 2016 Clovis-Cumberland-Dalton Succession: The Effects of Shott, Michael J. Environmental Change and 2002 Sample Bias in the Distribution and Colonization Processes on Abundance of Midwestern Fluted Technological Organization. Bifaces. Midcontinental Journal of PaleoAmerica (in press). Archaeology 27:89-123. 2013 Human Colonization and Late White, Andrew A. Pleistocene Lithic Industries of the 2006 A Model of Paleoindian Hafted Americas. Quaternary International Biface Chronology in Northeastern 285:150-160. Indiana. Archaeology of Eastern North America 43:29-59. Smallwood, Ashley M. 2012 Clovis Technology and Settlement Jesse W. Tune in the American Southeast Using Department of Anthropology Fort Lewis College Biface Analysis to Evaluate Durango, Colorado 81301 Dispersal Models. American [email protected] Antiquity 77(4):689-713.

Smallwood, Ashley M., Thomas A. Jennings, David G. Anderson, and Jerald Ledbetter 2015 Testing for Evidence of Paleoindian Responses to Environmental Changes during the Younger Dryas Chronozone in Georgia. Southeastern Archaeology 34(1):23-45.

Tankersley, Kenneth B. 1990 Late Pleistocene Lithic Exploitation in the Midwest and Midsouth: Indiana, Ohio, and Kentucky. In Early Paleoindian Economies of Eastern North America, edited by K. B. Tankersley and B. L. Isaac, pp. 259-299. Research in Economic Anthropology

41 QUANTIFYING REGIONAL VARIATION IN TERMINAL PLEISTOCENE ASSEMBLAGES FROM THE LOWER TENNESSEE RIVER VALLEY USING CHERT SOURCING

Ryan M. Parish and Adam Finn

The study examines the distribution of Terminal Pleistocene/Early Holocene groups in the Lower Tennessee River Valley by chert resource selection. The source data obtained via provenance analysis of 349 Middle/Late Paleoindian and Early Archaic bifaces potentially provides a method to differentiate hunter-gatherer groups through resource selection decisions. Periodic aggregation of Late Paleoindian groups is tentatively offered as an explanation for ‘exotic’ chert resources found within eight site assemblages. Analysis of the undocumented Late Paleoindian component of the Jim Parris collection compliments John Broster’s legacy of collaboration. John’s pioneering work with avocationalists in recording the spatial distribution of Paleoindian sites inspires a new generation of researchers.

The archaeological record rarely decisions. permits the analysis of individual social The transition from the late units segregated both spatially and Pleistocene to early Holocene temporally due to mixing of multiple environments and the possible cultural cultural components, taphonomic response to this climatic shift in the conditions, limitations of radiometric Southeastern United States has left dating techniques, and a slew of other researchers with a unique challenge in natural and anthropogenic variables. One interpreting the archaeological record of method in which archaeologists attempt to the region. The Western Valley or the delineate cultural units in the prehistoric Lower Tennessee River Valley (LTRV) is material record is through the a unique setting to study cultural categorization of chronologically adaptation to climatic stabilization of the diagnostic stone tools. These stone tool early Holocene as the region lies along a ‘forms’ are thought to have been convergence zone between the Coastal influenced by environmental, Plain and Highland Rim biomes. However, technological, and social variables. The the LTRV in Hardin County, Tennessee knowledge of how to manufacture stone has received relatively little attention tools was certainly not biological, rather among professional archaeologists even lithic technology was a learned skill and though it is known by the avocational therefore inherently embodies cultural community to contain extensive cultural traditions. The current study seeks to deposits. identify distinct Terminal Pleistocene/Early Prior to the documentation of the Jim Holocene (TPEH) culture groups within Parris collection there existed a gap in the the Lower Tennessee River Valley by known archaeological record of the utilizing standard biface typologies and Paleoindian to Early Archaic periods in chert sourcing. The objective is to present the LTRV within Hardin and the a model by which researchers can define surrounding counties. The middle section unique hunter-gatherer groups by form of the Tennessee River and Pickwick and their lithic resource selection Reservoir immediately to the south in

42 Chert Sourcing

northern Alabama has undergone county in the state that is dissected and extensive archaeological testing and not bordered by the Tennessee River. contains significant Paleoindian and Early Archaic documented deposits (Cambron Terminal Pleistocene Hunter-Gatherers 1956, 1958; Cambron and Hulse 1960; of the Southeast DeJarnette et al. 1962; Driskell 1996; Futato 1996; Hubbert 1989; Hulse and We admittedly know little about Wright 1989; Meyer 1995; Sherwood et al. Paleoindian lifeways in the Southeast. 2004; Soday 1954). Similarly, the section Much of what we do know comes from a of the valley near the Tennessee and few well-documented sites such as Dust Kentucky border to its confluence with the Cave, Carson-Conn-Short, Topper, and Ohio River has enjoyed extensive surveys Hardaway. Models for Paleoindian (Broster and Norton 1996; Freeman et al. behavior were once reliant upon those 1996; Rolingson 1964; Rolingson and derived from adjacent regions west of the Schwartz 1966). The Parris collection of Mississippi River and the Northeast. The artifacts, particularly the Paleoindian and comparative datasets from these regions Early Archaic assemblage, is an were referenced to describe Paleoindian invaluable asset for the archaeological groups of the Southeast as highly mobile community and fills a large gap in our hunter-gatherers either tethered to high knowledge of prehistoric occupation in the quality lithic sources or practicing LTRV. embedded procurement strategies (Daniel A significant portion of the data 2001; Gardner 1983, 1989; Goodyear pertaining to Paleoindian research in the 1979; Goodyear et al. 1989). The visual Southeast has stemmed from surface assignment of chert artifacts to material collected lithic artifacts and deflated type localities was used as a proxy to multicomponent sites many of which hypothesize large territory ranges often reside in private collections. Furthermore, extending over hundreds of kilometers, a since the region is lacking in stratified view greatly influenced by those Paleoindian age deposits, with some constructed in adjacent regions. Recently noted exceptions, a majority of the areas this view has been refined (Anderson chronology stems from typologies 1996; Anderson et al. 2015) to include a established by researchers working in far regional scale ‘place-oriented’ model of flung regions. exploitation of resource rich river valleys. For over twenty years the Parris family The river valleys are currently thought surface collected artifacts from various to have been ‘staging areas’ during earlier locales in Hardin County and maintained phases of colonization in the region detailed maps and notes of their finds. (Anderson 1990; Smallwood 2012). Later The Parris collection is primarily Paleoindian hunter-gatherer groups are comprised of lithic materials viewed as utilizing smaller territory ranges encompassing nearly every cultural and centered on river valleys and adjacent temporal period from Paleoindian to uplands. This model has been Mississippian. The Parris' collecting constructed in part from the distribution of activities were mainly focused along the diagnostic artifacts and chert type floodplains and drainages of the identifications. Dense concentrations of Tennessee River within Hardin County. diagnostic Paleoindian artifacts along both Hardin is unique because it is the only the Cumberland and Middle Tennessee

43 Tennessee Archaeology 8(1-2) Summer 2016

River Valleys may demonstrate that these and beech trees flourish in the well- areas were preferentially exploited during drained saprolitic soils and cherty coarse the terminal Pleistocene. Environmental gravels. reconstructions show that riverine Cryptocrystalline stone sources environments in the Southeast would utilized prehistorically are found in have provided favorable conditions for abundance within the Highland Rim. human occupation possibly earlier than Major chert bearing formations include the adjacent areas (Delcourt and Delcourt Devonian aged Camden formation, 1985; Hollenbach 2009). A paucity of Silurian aged Brassfield formation, the Paleoindian sites along the LTRV, where Mississippian aged Fort Payne, Warsaw, the river turns northward from the St Louis, and Ste. Genevieve limestone Pickwick Reservoir, previously gave formations. Additional sources of tool researchers an inaccurate view of stone materials include chert gravel and Paleoindian settlement. The gap in site cobble deposits within the Cretaceous distribution may once have been aged Tuscaloosa and Upland Complex interpreted as demarcating a northern (formally Lafayette) gravel formations. A macro-band centered on the Cumberland wide variation of color, texture, and size River and a second macro-band along the exists in these gravels including honey Middle Tennessee River of Northern colored Fort Payne and white Camden Alabama. However, the Paleoindian chert. Tertiary deposits of Fort Payne portion of the Parris collection fills this (Horse Creek/Pickwick) also may take on data gap and represents a significant crimson red, yellow, and black to grey contribution to our understanding of staining within the iron oxide laden human settlement/subsistence during the Cretaceous sands. TPEH. Methodology Environmental Setting Archaeological Sampling The lower section of the Tennessee River diverts northward from its easterly A total of 519 Paleoindian and early route in Lauderdale and Colbert counties, Holocene diagnostic projectile point/ Alabama. Here the river crosses the Fall knives from 35 sites was analyzed. Only line dropping in elevation at Muscle those sites (n=8) containing eight or more Shoals prior to the Mississippi, Alabama diagnostic bifaces are included in the and Tennessee border. Flowing current study so that potential patterning northward beyond the Pickwick Reservoir, in chert source might be discerned within the Tennessee River flanks the western an adequate site sample. Heavily edge of the Highland Rim physiographic reworked samples from these eight sites province prior to empting into the Ohio numbering 101 were not included in the River near Paducah, Kentucky. The LTRV analysis and await additional quantitative flows between the Coastal Plain type identification methods. Therefore, physiographic province of Western data from 349 near complete diagnostic Tennessee and the Highland Rim of bifaces collected from sites 40HR370, Middle Tennessee draining portions of 40HR15, 40HR381, 40HR456, 40HR458, over 12 counties. A mixed deciduous 40HR383, 40HR395, and 40PY308 are hardwood forest consisting of hickory, oak presented in Table 1.

44 Chert Sourcing

Table 1. Bifaces Analyzed in the Study Presented by Site and by Typological Classification.

45 Tennessee Archaeology 8(1-2) Summer 2016

The original numbering system used deposits. The majority of chert samples by Jim Parris is retained and listed consist of materials coming from respectively as D-8, H-69, H-16s, H-100, Mississippian aged carbonate formations H-101, H-69, H-28, and P-2 for future outcropping along the Highland Rim researchers working with the collection. physiographic province as well as the These site assemblages are multi- Interior Low Plateau and Valley and Ridge component localities; however, the (Figure 1). Paleoindian component is a significant portion at each. Late Paleoindian Projectile Point Typology projectile points were analyzed including Quad, Beaver Lake, Greenbrier and Projectile points included in the study Dalton variants. Also analyzed were a displayed light to moderate reworking and small sample of Middle Paleoindian and indicated distinct macroscopically Early Archaic types. identifiable diagnostic attributes. Typological assignments were based Chert Sampling primarily on Justice (1987) and Cambron and Hulse (1975) point type Geologic chert samples consist of 30 classifications. Heavily reworked points specimens from each of 76 diagnostic of the Late Paleoindian/Early deposits/outcrops for a total of 2,280 Archaic transition were not included as samples. Major material types/variants specific typologies could not be represented in the chert sample database confidently assigned. include Ste. Genevieve/Monteagle (Wyandotte), Upper St. Louis (Cobden, Chert Sourcing (VNIR and FTIR Kentucky Blue), Lower St. Louis (Dover), reflectance spectroscopy) Fort Payne (Buffalo River [Black/Tan], Bullseye), Tuscaloosa gravel (Horse Reflectance spectroscopy techniques Creek/Pickwick), Bangor, and Burlington. non-destructively record the interactions These major chert types were collected of matter and electromagnetic radiation at from Missouri, Illinois, Kentucky, both the atomic and molecular scales. Tennessee, Mississippi, Alabama and Reflectance spectroscopy is not a Georgia. Material types not currently quantitative geochemical technique where included in the study that may potentially macro, trace, and rare earth elements are affect source results are Upland Complex identified and counted. The data produced gravels, Camden and Brassfield chert. is in the form of reflectance intensity data. Recorded prehistoric quarry sites were Peaks or spectral features are indicative targeted for sampling first followed by of sample specific atomic and molecular primary procurement sites, both ancient configuration that absorbs a portion of the and modern in situ outcrops and alluvial incident electromagnetic radiation at a deposits. Samples were obtained across unique wavelength. Each absorption peak both the vertical and horizontal extent of is indicative of atomic and molecular the deposit utilizing a judgmental random composition. In chert the absorption selection method. Geologic formation peaks relate to micro-mineral groups or provenience was recorded by referencing impurities within the quartz matrix. USGS geologic quadrangle maps for in However, the dominant spectral features situ outcrops, residual and alluvial in chert are related to the silica (SiO2)

46 Chert Sourcing

FIGURE 1. Locations of analyzed chert deposits. molecule. Small impurities alter the shape as SiO2 are visible as vibrational features. of the silica features producing potentially Both the VNIR and FTIR datasets can be diagnostic features related to the mended together creating a composite diagenetic processes affecting a particular spectrum of a chert sample consisting of deposit within a particular region. 4,018 reflectance values each potentially Two complimentary reflectance diagnostic of parent formation and spectroscopy techniques are utilized in deposit. The chert samples comprising the study. The first, Visible Near-infrared the database are grouped by type (i.e. (VNIR) reflectance spectroscopy, records geologic formation) and by deposit. the interaction of matter with the visible Spectral data was collected from an and near-infrared portions of the interior surface of each of the chert electromagnetic spectrum. Absorption geologic samples and processed in order features are primarily related to atomic to highlight spectral slope changes and configuration, specifically outer valence eliminate noise. The resulting spectra electron fields. Fourier Transform Infrared were compiled into a chert spectral (FTIR) reflectance spectroscopy records database or spectral library of samples the interaction of matter with the middle with known geologic provenience. Two infrared portion of the electromagnetic accuracy tests were conducted within the spectrum. Dipole bonded molecules such chert database to investigate the ability of

47 Tennessee Archaeology 8(1-2) Summer 2016

the spectral data to characterize the chert collectively reported together. Early samples by parent formation and by Archaic types analyzed include Stilwell, deposit. The accuracy tests randomly Pine Tree, and LeCroy. The majority of selected a 10% sample and treated these types analyzed were those diagnostic of as having unknown provenience. the Late Paleoindian with Greenbriers Discriminant function analysis was comprising the largest component (Table selected as the multi-variant statistical 1). The small number of Middle method utilized to assign unknown Paleoindian (Clovis and Cumberland) and samples to known deposits. The accuracy Early Archaic types preclude definitive of the combined reflectance spectroscopy diachronic chert source analysis but the techniques was assessed by the ability to results are reported here to encourage correctly assign the 10% ‘unknown’ future research. samples back to their known formations and deposits. The internal accuracy tests Chert provenance consistently returned results of 99% correct assignment for both parent The results of the chert provenance formation and deposit provenience. data are organized into two analytical Spectra were collected from the 519 groups. The initial analysis sourced the bifaces non-destructively. The results of projectile points by chert type or in other 349 of these are reported below words to a geologic formation. This inter- representing diagnostic types from eight formational provenance data describes sites. A series of three measurements the range of chert material types within were taken per artifact and later averaged the eight site assemblages. At both sites a in order to provide a comprehensive reliance on Fort Payne chert dominates all spectral characterization of each other material types (Figure 2a, 2b; Table specimen. Where recent damage was 1). A biface not identified as Fort Payne present exposing the interior surface, a include a Pine Tree which is characterized spectral reading was taken for as Ste. Genevieve. Other than this single comparison with the surficial point, a strong preference for Fort Payne measurement. Artifact spectra were resources at all eight sites is apparent. processed in the same manner as the The large geologic occurrence and geologic samples again in order to availability of Fort Payne chert spanning standardize measurements, eliminate six states and 600 linear kilometers noise and highlight spectral features. makes finer spatial source determinations necessary. Results A second statistical analysis was conducted upon all projectile points typed Paleoindian Projectile Point Types as Fort Payne to determine the source region(s) and identify any patterns. The All sites exhibit a range of Paleoindian intra-formation analysis is designed to point types spanning the TPEH (Table 1). source the projectile points to specific Diagnostic types analyzed include Clovis, areas within the formation. The 40 Cumberland, Beaver Lake, Quad, Dalton sampled Fort Payne deposits located from and Greenbrier. Dalton variants (Dalton/ the southern tip of Illinois to the Greenbrier, Dalton/Colbert, Dalton/ northwestern corner of Georgia including Nuchols) were identified but are deposits in central Kentucky, central

48 Chert Sourcing

FIGURE 2. a (upper): Discriminant Function scatter plot showing the characterization of diagnostic bifaces to Fort Payne chert samples; b (lower): Three dimensional Discriminant Function scatter plot showing the characterization of diagnostic bifaces to Fort Payne chert samples. Dover group is located behind the Fort Payne.

Tennessee, the northeastern corner of typed as Fort Payne chert. Mississippi and northern Alabama were The discriminant function analysis was used to refine the spatial resolution of the rerun only including each of the 40 provenance data for only those bifaces sampled Fort Payne Deposits as potential

49 Tennessee Archaeology 8(1-2) Summer 2016

source deposits. Three Fort Payne source application of reflectance spectroscopy regions were exclusively identified from instrumentation to chert provenance this intra-formation provenance assay. research studies demonstrates the rich The Fort Payne source regions include cultural data it provides. The benefits of the Western Highland Rim of Tennessee, reflectance spectroscopy include its Northern Alabama, and Southern Illinois relatively low cost, speed, accuracy and (Table 1). The majority of all bifaces non-destructive potential. However, any (51%, n=179) were manufactured from chert sourcing technique is arguably only Fort Payne deposits found along the as good as the comparison database Western Highland Rim of Tennessee, within which unknown artifacts are being specifically those deposits located in characterized. The chert type database Houston, Humphreys, Hickman, and used in the current study consists of 2,280 minor amounts from Henry and Wayne samples obtained from 76 chert deposits counties. Fort Payne deposits from and collectively represents seven material northern Alabama accounted for a total of types. Additional sampling of local chert 16% (n=57). The third source region is types including Brassfield, Camden and located at a greater distance from the Upland Complex gravel sources may study area from deposits located at the influence future results. southern tip of Illinois contributing Viewing the results from a approximately 32% (n=112) of all Fort technological perspective includes Payne represented in the diagnostic recognition of the potential perils of outer biface assemblages (Table 1). ‘patina’ surface analysis on archaeological The results of the spatially refined materials. Geochemical studies provenance analysis broadly demonstrate demonstrate that chemical alterations are a reliance upon ‘local’ deposits of Fort present upon the patina surface of some Payne chert located in close proximity to chert types (Gauthier et al. 2012). the sites. The use of local material by Preliminary experimentation by the lead Middle Paleoindian inhabitants is author shows that patina may in part be potentially demonstrated by the source due to the increased angular micro- data upon Clovis and Cumberland bifaces surface topography occurring on chert (Table 1). The occurrence of ‘non-local’ that has undergone mechanical and Fort Payne materials from Southern chemical weathering of mineral grains Illinois do not appear in the bifaces until (Parish 2013). The elimination of noise- those diagnostic of terminal Late dominated regions in the visible portion of Paleoindian/Early Archaic times. The the electromagnetic spectrum appears to exception to this trend is at site 40HR381 alleviate patina variations in the spectral where Fort Payne material from Southern data. Dual measurements obtained upon Illinois comprises the majority of stone the outer and inner surfaces of artifacts tool material. exhibiting modern edge damage illustrate the minor differences that patina formation Discussion has on spectral measurements (Figure 3). However, additional studies are being The provenance results presented conducted to explore these effects. here need to be contextualized from a Continuing research and development of methodological, technological, theoretical outer surface patina effects is crucial if and cultural perspective. The continuing reflectance spectroscopy is to be used as

50 Chert Sourcing

FIGURE 3. Discriminant function analysis depicting the spectral variance of interior (unweathered) vs. exterior (patina) surface measurements taken upon diagnostic bifaces with patina and modern damage exposing inner surfaces. a non-destructive chert provenance diagnostic bifaces found at sites from this technique. section of the LTRV. Currently, no known The provenance data should be sources of high quality lithic material exist grounded in theory if human behavioral in the immediate (50 km radius) area data is to be gleaned from the study. It is though continued evaluation of alluvial hypothesized that TPEH groups in the sources may change this observation. Tennessee and Cumberland River Valleys The Yellow Creek drainage basin south of were settling into resource rich pockets, the region in Mississippi may be an exploiting local resources, and decreasing exception (Johnson 1981). range mobility. The large 40HR370 site, A greater knowledge of the chert possibly located on a terrace of the gravel deposits within the LTRV will Tennessee River, could represent a almost certainly clarify our understanding seasonal or periodically occupied of locally available tool stone resources. residential base camp from which As noted by other researchers (Anderson logistical forays into the Coastal Plain and et al. 2010), groups during the TPEH Highland Rim could have been organized appear to rely more upon local chert (Binford 1980). However, Gardner’s deposits. If local exploitation of (1983, 1989) tethered to high quality tool Tuscaloosa/Fort Payne gravels did occur stone model does not seem to explain the in the region than analysis of the debitage relatively large quantities of Paleoindian and discarded bifaces would give us clues

51 Tennessee Archaeology 8(1-2) Summer 2016

regarding initial reduction of river cobbles. Indirect acquisition via down the line Jim Parris on multiple times denied the exchange, maintaining social networks, appearance of large amounts of primary gifting, etc. are another explanation for the reduction material at these sites. A presence of exotic Fort Payne at the cursory examination of his collection TPEH sites along the LTRV. Indirect including the unprovenienced portion acquisition is difficult to visualize and reveals the paucity of early stage bifaces often requires additional datasets such as and cores, whether this is a function of in depth lithic analysis of site collector bias is currently unclear. If in fact assemblages (Meltzer 1989). However, Gardner’s “place-oriented” model is an the inhabitants of the LTRV had access to adequate explanation of site location, as high quality deposits of chert along the adopted by other researchers in the Highland Rim of both Tennessee and Southeast (Anderson 1996; Daniel 2001), Alabama albeit possibly not in the then TPEH inhabitants may have been immediate vicinity. Indirect acquisition there for resources other than chert may not adequately explain the chert acquisition. provenance patterning revealed as Traditionally, two main models are relatively large amounts of Fort Payne proposed to explain the presence of from Illinois are identified at each site. In exotic chert materials in archaeological fact at 40HR381, Southern Illinois Fort assemblages, direct and indirect Payne comprises nearly all of the acquisition. Highly (residential) mobile diagnostic biface assemblage. If indirect foragers is the commonly accepted view acquisition was the mechanism for the of both early and middle Paleoindian Illinois Fort Payne than one might expect groups in adjacent regions. As mentioned smaller numbers of these exotics. previously, this view is challenged in the Direct acquisition during periodic Southeast due to the spatial distribution of subsistence rounds and indirect diagnostic bifaces and the prevalence of acquisition may not account for the data ubiquitous high quality chert resources. revealed in this study but the Movements of Late Paleoindian congregation of distant groups in order to (Greenbrier and Dalton) groups and/or exchange information is worth macro-bands are one possible consideration. The series of TPEH sites explanation for the presence of Fort along this portion of the LTRV may Payne chert from Northern Alabama and represent occasional congregation areas Southern Illinois. A seasonal or periodic for three macro-bands centered locally territory range encompassing over 30,000 along the Western Highland Rim, square kilometers narrowly defined from Northern Alabama and the confluence of the confluence of the Ohio and Mississippi the Ohio and Mississippi Rivers (Figure Rivers to the Middle Tennessee River 4). Periodic aggregations of distant hunter Valley in Northern Alabama, pushes the gatherer groups, termed fandangos by boundaries for historically documented Steward through his observations of hunter gatherer societies (Binford 1983; Reese River Valley Shoshone, would Kelly 2007; Steward 1938). Even with the have been important to peoples adjusting use of watercraft, the question remains to shifting resources. Exchange of what would draw LPEH groups northward, information, mates, and materials is certainly not the relatively small deposits documented ethnographically though in of Fort Payne (Elco) chert. differing environmental settings (Binford

52 Chert Sourcing

FIGURE 4. Three main Fort Payne source regions identified in the study possibly signifying periodic aggregations of three Late Paleoindian/Early Archaic macro- bands. 1983; Steward 1938; Thomas 1973). visual identifications upon diagnostic Though a tenuous explanation, periodic bifaces by other researchers suggest aggregation between macro-bands along otherwise (Koldehoff 2013; Koldehoff and this section of the LTRV may also account Loebel 2009; Lopinot and Butler 1981). for the overabundance of Southern Illinois Additionally, a complete collection and Fort Payne chert at sites 40HR381 and analysis of any primary reduction debitage 40PY308. These sites being repeated from these sites may give clues to primary camp locations for those northern acquisition mechanisms. Finally, the bands. Only continued analysis from vertical integrity of the sites needs to be methodological, technological, theoretical assessed and the potential of intact and cultural perspectives can clarify the deposits needs to be evaluated. Through results presented here. future research, a greater understanding In order to test the preliminary results of the complex and variable record of generated in the current study, TPEH TPEH peoples may be realized. assemblages further to the north require chert provenance analysis. Subsequent results may substantiate the use of “Elco” Fort Payne chert by TPEH groups though

53 Tennessee Archaeology 8(1-2) Summer 2016

Conclusions extending back from European contact to the earliest inhabitants of Tennessee. The The provenance results of 349 current study hopefully exemplifies John Terminal Pleistocene/Early Holocene Broster’s legacy of working with diagnostic bifaces from eight sites along avocational archaeologists and collectors the Lower Tennessee River Valley in piecing together an understanding of illustrate the almost exclusive use of Fort life at the Holocene boundary. Payne chert resources. Deposits located along the Western Highland Rim Acknowledgements. The authors wish to collectively represent the majority of acknowledge the late Jim Parris and his family for their gracious hospitality on many occasions, material as one would expect for Late enthusiasm, and love for archaeology and life. Paleoindian regional foragers but a Also, we acknowledge the legacy of John Broster significant portion of the diagnostic which continues to inspire a new generation of bifaces are sourced to deposits in archaeologists. All errors, oversights, and Northern Alabama and Southern Illinois. overreaching hypotheses found within are the authors’ sole responsibility. An earlier version of Explanations for this include this was presented at the 2013 SEAC meeting in methodological/technological issues Tampa, Florida November 7-9, 2013. related to sampling and/or analysis upon outer artifact patina surfaces, direct References procurement along large territorial ranges, indirect acquisition and periodic Anderson, David G. aggregations. Currently given the large 1990 The Paleoindian Colonization of chert type database and preliminary Eastern North America: A View studies regarding the effects of outer from the Southeastern United artifact surface analysis, methodological States. In Early Paleoindian and technological concerns may be Economies of Eastern North deferred. Both the size of the proposed America, edited by Kenneth band-group territory range and number of Tankersley and Barry Isaac, pp. exotic Fort Payne materials encountered 163–216. Research in Economic per site is taken as evidence discounting Anthropology Supplement 5. direct procurement by a single group and 1996 Models of Paleoindian and Early indirect acquisition. Therefore, it is Archaic Settlement in the proposed that evidence for periodic Southeastern United States. In The aggregation events along the LTRV Paleoindian and Early Archaic during the TPEH period exists. Southeast, edited by D. G. Apart from the interpretations Anderson, and K. E. Sassaman, regarding site function and distributions of 29–57. Tuscaloosa: University of chert materials, it is apparent that this Alabama Press. relatively small portion of the well provenienced Jim Parris collection Anderson, David G., Ashley M. significantly impacts our understanding of Smallwood and D. Shane Miller prehistoric life along the southern reaches 2015 Pleistocene Human Settlement in of the Lower Tennessee River Valley. the Southeastern United States: Subsequent studies that incorporate data Current Evidence and Future from the Parris collection will greatly Directions. PaleoAmerica 1: 7-51. clarify our understanding of lifeways

54 Chert Sourcing

Anderson, David G., D. Shane Miller, S. J. Daniel, I. R., Jr. Yerka, J. C. Gillam, E. N. 2001 Stone Raw Material Availability and Johanson, D. T. Anderson, A. C. Early Archaic Settlement in the Goodyear, and A. M. Smallwood Southeastern United States. 2010 PIDBA (Paleoindian Database of American Antiquity 66: 237–265. the Americas) 2010: Current Status and Findings. Archaeology of DeJarnette, David L., Edward B. Kurjack, Eastern North America 38: 63–90. and James W. Cambron 1962 Excavations at the Stanfield- Binford, Louis R. Worley Bluff Shelter. Journal of 1980 Willow Smoke and Dogs’ Tails: Alabama Archaeology 8(1-2):1- Hunter-gatherer Settlement 124. Systems and Archaeological Site Formation. American Antiquity 45: Delcourt, H. R., and P. A. Delcourt 1–17. 1985 Quaternary Palynology and 1983 In Pursuit of the Past: Decoding the Vegetational History of the Archaeological Record. Thames Southeastern United States. In and Hudson, New York, New York. Pollen Records of Late-Quaternary North American Sediments, edited Broster, John B., and Mark R. Norton by V. M. Bryant, and R. G. 1996 Recent Paleoindian Research in Holloway, 1–37. Dallas: American Tennessee. In The Paleoindian Association of Stratigraphic and Early Archaic Southeast, Palynologists Foundation. edited by David G. Anderson and Kenneth E. Sassaman, pp. 288- Driskell, Boyce N. 297. The University of Alabama 1996 Stratified Late Pleistocene and Press, Tuscaloosa. Early Holocene Deposits at Dust Cave, Northwestern Alabama. In Cambron, James W. The Paleoindian and Early Archaic 1956 The Pine Tree Site -- a Paleo- Southeast, edited by David G. Indian Habitation Locality. Anderson and Kenneth E. Tennessee Archaeologist 12(2):1- Sassaman, pp. 315-330. The 10. University of Alabama Press, 1958 Paleo Points from the Pine Tree Tuscaloosa. Site. Tennessee Archaeologist 14(2):80-84. Freeman, Andrea K. L., Edward E. Smith, Jr., and Kenneth B. Tankersley Cambron, James M. and David C. Hulse 1996 A Stone's Throw from Kimmswick: 1960 An Excavation on the Quad Site. Clovis Period Research in Tennessee Archaeologist 16(1):14- Kentucky. In The Paleoindian and 26. Early Archaic Southeast, edited by 1975 Handbook of Alabama David G. Anderson and Kenneth E. Archaeology; Part I Point Types. Sassaman, pp. 385-403. The The Archaeological Research University of Alabama Press, Association of Alabama Inc., Tuscaloosa. Moundville.

55 Tennessee Archaeology 8(1-2) Summer 2016

Futato, Eugene M. Goodyear, A. C., III, J. L. Michie, and T. 1996 A Synopsis of Paleoindian and Charles Early Archaic Research in 1989 The Earliest South Carolinians: Alabama. In The Paleoindian and The Paleoindian Occupation of Early Archaic Southeast, edited by South Carolina. Occasional Papers David G. Anderson and Kenneth E. 2. Columbia: Archaeological Sassaman, pp. 298-314. The Society of South Carolina. University of Alabama Press, Tuscaloosa. Hollenbach, K. D. 2009 Foraging in the Tennessee River Gardner, W. M. Valley 12,500 to 8,000 Years Ago. 1983 Stop Me if You’ve Heard this One University of Alabama Press, Before: The Flint Run Paleoindian Tuscaloosa. Complex Revisited. Archaeology of Eastern North America 11: 49–59. Hubbert, Charles M. 1989 An Examination of Cultural Change 1989 Paleo-Indian Settlement in the in the Late Pleistocene and Early Middle Tennessee Valley: Holocene (circa 9200 to 6800 Ruminations from the Quad Paleo- B.C.). In Paleoindian Research in Indian Locale. Tennessee Virginia: A Synthesis, edited by J. Anthropologist 14:148-164. M. Wittkofski, and T. R. Reinhart, 5–51. Special Publication 19. Hulse, David C., and Joe L. Wright Richmond: Archeological Society of 1989 The Pine Tree-Quad-Old Slough Virginia. Complex. Tennessee Anthropologist 14:102-147. Gauthier, Gilles, Adrian L. Burke, and Mathieu Leclerc Johnson, Jay K. 2012 Assessing XRF for the 1981 Lithic Procurement and Utilization Geochemical Characterization of Trajectories: Analysis, Yellow Radiolarian Chert Artifacts from Creek Nuclear Power Plant Site, Northeastern North America. Tishomingo County, Mississippi Journal of Archaeological Science Volume 2. Archaeological Papers 39 (7): 2436-2451. of the Center for Archaeological Research, University of Mississippi, Goodyear, A. C., III. University. 1979 A Hypothesis for the Use of Cryptocrystalline Raw Materials Justice, Noel D. among Paleoindian Groups of 1987 Stone Age Spear and Arrow Points North America. In Eastern of the Midcontinental and Eastern Paleoindian Lithic Resource Use, United States. Indiana University edited by C. J. Ellis, and J. C. Press, Bloomington. Lothrop, 1–9. Boulder: Westview Press. Kelly, Robert L. 2007 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Percheron Press, New York, New

56 Chert Sourcing

York. Weaponry. Unpublished Ph.D. dissertation, Department of Earth Koldehoff, Brad H. Sciences, University of Memphis. 2013 Paleoindian and Archaic Settlement and Lithic Procurement Rolingson, Martha Ann in the Illinois Uplands. Technical 1964 Paleo-Indian Culture in Kentucky: Report No. 148. Illinois State A Study Based on Projectile Points. Archaeological Survey, Urbana- Studies in Anthropology, No. 2. Champaign. University of Kentucky Press, Lexington. Koldehoff, Brad H. and Thomas J. Loebel 2009 Clovis and Dalton: Unbounded and Rolingson, Martha Ann, and Douglas W. bounded systems in the Schwartz Midcontinent of North America. In 1966 Late Paleo-Indian and Early Lithic Materials and Paleolithic Archaic Manifestations in Western Societies, edited by Brian Adams Kentucky. Studies in Anthropology and Brooke S. Blades, 270–286. No. 3. University of Kentucky Oxford: Blackwell Publishing Ltd. Press, Lexington.

Lopinot, Neal H., and Brian M. Butler Sherwood, Sarah C., Boyce N. Driskell, 1981 Archaeological Assessment of Asa R. Randall, and Scott C. Exchange Lands in Alexander Meeks County, Illinois Shawnee National 2004 Chronology and Stratigraphy at Forest. Research Paper No. 28. Dust Cave, Alabama. American Center for Archaeological Antiquity 69:533-554. Investigations, Southern Illinois University, Carbondale. Smallwood, Ashley M. 2012 Clovis Technology and Settlement Meltzer, David J. in the American Southeast: Using 1989 Why Don’t We Know When the Biface Analysis to Evaluate First People Came to North Dispersal Models. American America? American Antiquity Antiquity 77(4):689-713. 54:471–490. Soday, Frank J. Meyer, Catherine C. 1954 The Quad Site, a Paleo-Indian 1995 Cultural Resources in the Pickwick Village in Northern Alabama. Reservoir. Report of Investigations Tennessee Archaeologist 10(1):1- 75. The University of Alabama, 20. Alabama Museum of Natural History, Division of Archaeology, Steward, Julian H. Moundville. 1938 Basin Plateau Aboriginal Sociopolitical Groups. Bureau of Parish, Ryan M. American Ethnology Bulletin 120. 2013 The Application of Reflectance Smithsonian Institution, Spectroscopy to Chert Provenance Washington, D.C. of Mississippian Symbolic

57 Tennessee Archaeology 8(1-2) Summer 2016

Thomas, David H. 1973 An Empirical Test for Steward’s Model of Great Basin Settlement Patterns. American Antiquity 38:155-176.

Ryan M. Parish University of Memphis 125 Johnson Hall Memphis TN 38152

Adam Finn University of Memphis 110 Johnson Hall Memphis TN 38152

58 A PRELIMINARY REPORT ON THE LATE PLEISTOCENE AND EARLY HOLOCENE ARCHAEOLOGY OF ROCK CREEK MORTAR SHELTER, UPPER CUMBERLAND PLATEAU, TENNESSEE

Jay Franklin, Maureen Hays, Frédéric Surmely, Ilaria Patania, Lucinda Langston, and Travis Bow

Rock Creek Mortar Shelter (40PT209), in Pickett State Forest on the Upper Cumberland Plateau of Tennessee, possesses an intermittent 11,500 year occupation history. This history may be consistent with previous ideas of first colonization of upland rock shelter zones at the end of the Younger Dryas with significant climatic amelioration. However, culturally sterile deposits have yet to be encountered and the site may be older still. This work focuses on the late Pleistocene and early Holocene components, paying particular attention to unifacial, blade, and blade-like tool production and technology, use-wear analysis, and depositional history. Variability in blade production during the late Pleistocene deposits suggests residentially mobile family groups, and could also represent the colonizers’ struggles with adapting a blade tool technology to the locally abundant small, rounded Monteagle chert cobbles.

Three field seasons have been Big South Fork river valleys that represent completed at Rock Creek Mortar Shelter the ends of early migration routes as (40PT209) in Pickett State Forest on the noted by Lane and Anderson (2001). As Upper Cumberland Plateau (UCP). The the climate ameliorated beginning about archaeological deposits are comparatively 11,600 years ago, hardwood forest deep and may extend back into the communities migrated to higher Paleoindian period. Radiocarbon dates elevations. People began to exploit nut suggest this to have occurred perhaps as mast resources and associated game early as 12,500 years ago but certainly by animals. A seasonal round, or way of 11,500 years ago. The shelter lies at what doing, was established. This way of doing Lane and Anderson (2001) refer to as a things in the uplands set the tone for the migration terminus - the end of early next several millennia. migration routes. The early occupational Site 40PT209 is important because it history of the shelter is thus far consistent represents the first recorded in 20 years with Walthall’s (1998) ideas of first of work on the UCP with late Pleistocene colonization of upland rock shelter zones and early Holocene deposits that appear at the end of the Younger Dryas with intact and in relative stratigraphic position. significant climatic amelioration. The site The previous work includes intensive was intermittently occupied over the surveys of the East Fork Obey gorge course of the next 11,500 years until (Franklin 2002, 2006), Pogue Creek about AD 1000. Canyon State Natural Area (Langston The Upper Cumberland Plateau was 2013; Langston and Franklin 2011), and likely a very different environment 12,500 significant portions of Pickett State Forest years ago at the end of the Pleistocene (Langston et al. 2012). Culturally sterile during the Younger Dryas. Still, some layers have not been encountered at early pioneers ventured their way up on Rock Creek Mortar Shelter, thus it is the plateau, perhaps up the Wolf and/or possible that more ancient archaeological

59 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 1. Rock Creek Mortar Shelter looking north. levels could be present. The shelter and fine to coarse-grained (Phillips et al. represents a great opportunity to examine 2010:189). This formation ranges in cultural and technological change through thickness from about 46 to 67 meters time as Early Archaic through Late (about 150 – 220 feet) and is typically the Archaic and Woodland components are highest elevation formation in this area of present. This provides not only an the UCP. This high elevation rock shelter extremely rare opportunity to examine is one of tens of thousands in an upland colonizing exploitation in an upland region dominated by rugged terrain, region, but also allows an opportunity to precipitous stream gorges, and hundreds examine an 11,000-year prehistoric of kilometers of sandstone bluff lines. The cultural sequence at one place. shelter does not occur along a major river Rock Creek Mortar Shelter faces west as the closest stream is Rock Creek and lies at the northern extremity of a long approximately 400 meters due north. upland bluff line (Figure 1). The shelter is However, there is a 70-meter high bluff situated within the Rockcastle line that separates the rolling uplands Conglomerate at an elevation of 464m where the shelter sits and Rock Creek amsl (about 1520 feet). The Rockcastle is (Figure 2). a conglomeratic sandstone (consisting of Rock Creek Mortar Shelter has not densely packed quartz pebbles) and been disturbed by vandals or artifact sandstone (with a few scattered rounded hunters, a rarity in a region with a 150- quartz pebbles), gray to brown in color, year history and tradition of artifact

60 Rock Creek Mortar Shelter

Rock Creek Mortar Shelter is apparently not the only early site in the Rock Creek drainage of the Little South Fork Cumberland River. Prismatic blades were recovered during survey of sites 40PT216 (open air ridge top) and 40PT241 (Hot Bluff Shelter) in 2012. Research has also recorded several additional pristine rock shelter sites, suggesting the drainage is a high value area to continue archaeological survey and seek out potential Paleoindian and Early Archaic sites. One issue we propose to address is Lane and Anderson’s (2001:92) proposition that the Upper Cumberland Plateau lies at the physiographic end of an early migration route. “There have been no research projects directed to exploring Paleoindian occupations in the area, and large-scale FIGURE 2. Rock Creek Mortar Shelter study area. professional excavation of hunting and looting (Des Jean and known sites is non-existent” Benthall 1994:139; Franklin 2002:5-6). (Lane and Anderson 2001:94). Therefore, The site is located near a primary road what little we do know may actually reflect and Pickett State Park, and its integrity is investigation bias rather than settlement very much in danger from illegal digging. preference on the part of Paleoindian Park personnel monitor the shelter, but it populations. This notion is not surprising is unlikely the site can be protected as “there is an understandable research indefinitely. A long term excavation bias against examining steeply sloping program is planned in conjunction with the terrain and this. . . might reduce the 2016 opening of the Pickett State Park likelihood of discovering Paleoindian Archaeology Museum and ETSU sites” (Lane and Anderson 2001:90). Archaeological Research Station. The site However, steeply sloping terrain often will be an integral part of public outreach includes rock shelters, and there are at the park. thousands that remain to be investigated

61 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 3. Rock Creek Mortar Shelter plan view, January 2014. on the Upper Cumberland Plateau. The dedicated to test excavations at the work at Rock Creek will anchor our shelter. Woodland and Late Archaic research program aimed at the late period artifacts were recovered within two Pleistocene and Early Holocene test units (Units 1 and 2; Figure 3) in good archaeology of the UCP. stratigraphic position. ETSU returned to the site in March History of Work at Rock Creek Mortar 2013 for further testing. In addition to Shelter continuing with the two open test units, a 3 x 1 meter test trench was established Tennessee State Parks ranger Travis perpendicular to the shelter’s dripline axis. Bow discovered Rock Creek Mortar In this trench, at about 85 cm below Shelter in early 2012. Two things drew his surface, we encountered Early Archaic attention: (1) the presence of a large artifacts. This discovery was the first time bedrock mortar hole in a back ledge, and Early Archaic materials had been (2) the site was pristine – something discovered by ETSU in stratified contexts previously noted as an uncommon on the UCP. Two specimens recovered at occurrence on the UCP. Jay Franklin 125 cm below surface (a double side visited the site with Bow in March 2012, scraper made on a blade, and a large and returned in January 2013 as part of biface with possible overshot flaking) East Tennessee State University’s favorably compared to Paleoindian (ETSU) inaugural winter session field artifacts, possibly Clovis (John Broster, school. The focus of the field school was personal communication, 2013). largely survey for new sites in Pickett A second winter season of archaeo- State Forest, but the last few days were logical testing was conducted in

62 Rock Creek Mortar Shelter

December 2013. The trench was Stratigraphy extended another three meters to beyond the drip line. Also, four additional test Test Units units were opened along the same east- west line as the two previously mentioned Stratigraphy in the (eventual) 7 x 1 units, in effect creating a trench running meter perpendicular trench appears parallel to the shelter’s axis (Figure 3). distinct from the test units (farther back in Numerous blades and blade fragments the shelter) that run parallel with the long were recovered along with a few axis of the shelter. In the test units (Figure diagnostic bifaces. Five AMS radiocarbon 4), Stratum 1 is a recent humus layer that assays date the late Pleistocene/early varies between 5-10 cm thick. Stratum 2 Holocene deposits at Rock Creek Mortar is pale yellow to white sand about 20 cm Shelter. thick in the south part of the shelter that A third field season was completed in pinches out moving northward, and winter 2014/2015. Several more units appears to be the result of weathering of a were opened, and the trench was sandstone ledge 2.5 meters above the extended further beyond the shelter drip surface. Stratum 2 has a horizontal extent line. Additional blades and blade-like of about two meters that is consistent with flakes were recovered, and the earliest the ledge. Woodland period artifacts were diagnostic bifaces originate from the Early recovered from the mid to lower levels of Archaic period. Eight additional AMS this layer. dates were obtained during Field Season Stratum 3 is a strong brown loamy 3, and all but one AMS date associate sand layer that yielded Woodland period these artifacts with the Early Archaic artifacts from the top level, and Terminal period. Archaic artifacts below that. Stratum 3 also yielded evidence of thin, poorly

FIGURE 4. East wall profile of test units, January 2014.

63 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 5. South wall profile of trench, January 2014. preserved, prepared baked clay surfaces. white sand that may correspond with Stratum 4 comprises a rich organic, Stratum 7 in the trench. A few non- very dark brown to black loamy sand that diagnostic artifacts were recovered from corresponds to the Late and Terminal this essentially sterile layer. Archaic periods. This layer varies in thickness and extent and may be better Trench interpreted as a midden rather than a lithostratigraphic unit. It could also be due The trench stratigraphy is different in part to differential accumulation of than described for the test units (Figure organics due to significant rock fall in the 5). The humus layer (Stratum 1) is visible, shelter. Baked clay surfaces were also but the yellowish white to white sand encountered in Stratum 4, with some (like Stratum 2 is not present (the trench is Feature 5) better preserved than those removed from the weathering ledge here). encountered in Stratum 3. The presence Below the humus is a 20-30 cm thick, of baked clay surfaces is consistent with strong brown to orange brown loamy sand the recovery of nutting stones in Stratum layer similar to Stratum 3. A large steatite 4. Elsewhere, these surfaces have been bowl fragment in Stratum 3 probably described as having been used to parch dates to the Terminal Archaic period. nuts (Homsey-Messer 2015). Below Stratum 3 is a 20-40 cm thick Below Stratum 4 is a brown sand layer layer of yellowish brown loamy sand defined as Stratum 8 (possibly the result defined as Stratum 5. Early Archaic of organics leaching from Stratum 4) artifacts were encountered toward the containing Middle and Early Archaic bottom of this layer. Below Stratum 5 is an period artifacts. Stratum 9 is composed of organic-rich layer that does not

64 Rock Creek Mortar Shelter correspond to the previously noted Stratum 4 midden observed in the test units. This layer (Stratum 6) appears to result from organic materials being trapped under a massive slab of sandstone breakdown as it trails off toward the drip line and away from the slab. Early Archaic artifacts were recovered in this layer. Below the organic-rich layer is approximately 25 cm of yellow sand (Stratum 7) that may correspond with the white sand layer (Stratum 9) in the excavation units farther back under the shelter. Stratum 7 FIGURE 6. Layer 6 composed of well sorted quartz sand yielded materials interpreted and disorthic Mn nodules coated in dusty clay, deposited by as transitional Paleoindian/ alluvial processes. Early Archaic. Excavations in the trench presence of an anthropogenic fire. were discontinued at 1.75 to 2.0 meters Stratum 6 will need to be sampled in below surface. A bucket auger revealed greater detail to clarify questions of another 62 cm of sediment below this provenience and anthropogenic input. level with more organic content and more The composition and arrangement of clay in the sand. However, bedrock has Stratum 7 (RCMS 15-5) suggests the yet to be reached. sediments are the result of in-place dissolution of eboulis and roof fall. Post- Micromorphology depositional processes in this layer are minimal. Manganese crusts and staining Micromorphological analyses are confirm water saturation for short periods ongoing, with some preliminary results of time, though there is no evidence of the from Strata 6 and 7 presented in this work sediment being waterlogged. Bioturbation {samples numbers RCMS 15-5 (Stratum in the form of plants and insects is not 7), 15-6 (Stratum 6/7 contact), and 15-7 visible, but more analyses are needed to (Stratum 6)}. Sample 15-7 (Stratum 6) is assess the overall impact of bioturbation composed of sorted quartz sand with on this layer. dusty clay coatings and rounded disorthic Preliminary micromorphological analy- manganese nodules (Figure 6). There is sis suggests that Stratum 7 is not a currently no stream nearby the shelter, cultural layer, as the artifacts and charcoal but the sediments in Stratum 6 have the present may be translocated from the appearance of being deposited by alluvial layer above. Formation of Stratum 6 is processes from a nearby stream or other more complex as it resulted from the body of water. There is no evidence in the combination of geogenic and sample of a living surface. Although anthropogenic inputs. However, no living charcoal fragments are observed, no ash surface or intact fire signatures were or burned bone is visible ruling out the observed, suggesting Stratum 6

65 Tennessee Archaeology 8(1-2) Summer 2016

represents ephemeral and intermittent visits to the site by the shelter’s early inhabitants. Cultural and technological associations may also be tenuous.

Chronology

Diagnostic artifacts recovered from Rock Creek Mortar Shelter indicate a long temporal span from the Early Archaic through Woodland periods. In addition there are 20 AMS radiocarbon dates from the site (calibrated using CALIB 7.0), and also two OSL (optically stimulated luminescence) dates from a Woodland ceramic sherd and a baked clay surface (Table 1).

Early Holocene/Late Pleistocene

Eight AMS determinations from wood charcoal date to the early Holocene and one other dates to the late Pleistocene. Six derive from the trench excavation, and FIGURE 7. Early Archaic bifaces, Rock Creek two derive from Features 4 and 6 in Mortar Shelter. Stratum 8 (Test Unit 2). Two assays from boundary at 11,260 cal BP. The previous the trench inside the drip line are 9073 three dates were obtained from charcoal and 9550 cal BP, (piece plots 164 and 66, found in spatial association with 18 respectively). Spatially associated prismatic and triangular blades and blade artifacts are Kirk Corner Notched and Lost fragments. Three bladelets were also Lake bifaces. Piece plots 193 (9594 cal recovered. Associated bifaces include a BP) and 198 (9598 cal BP) are from heavily reworked Greenbrier Dalton (see Features 6 and 4, respectively. Spatially Figure 7). Finally, an assay of 12,554 cal associated artifacts are a Hardaway Side- BP was obtained from Level 5 (piece plot Notched biface and a prismatic blade core 282) of the trench extension beyond the fragment from Stratum 8. A calibrated drip line. Spatially associated artifacts median date of 9269 cal BP was obtained include a blade core fragment and several from Level 4 (piece plot 273) in the trench blade-like flakes. extension beyond the drip line (Stratum Radiometric dates from Rock Creek 7). Other recovered Early Archaic bifaces Mortar Shelter suggest occupation from from the site include several bifurcated the late Pleistocene (Younger Dryas) specimens (Figure 7). Two dates precede through the late Holocene. However, 10,000 years ago at 10,623 and 10,760 Stratum 7 does not appear to be a cultural cal BP (piece plots 96 and 127). One layer, and Stratum 6 located directly assay from the trench (piece plot 166) above is problematic with sediments that straddles the Pleistocene/Holocene may be partially cultural but also alluvial.

66 Rock Creek Mortar Shelter

Table 1. Radiometric Dates from Rock Creek Mortar Shelter (40PT209).

Caution must be exercised when Most of the intentionally-produced considering the association of these blades (n=18, and a few bladelets) were charcoal samples with the archaeological made/prepared from unipolar cores. materials. The earliest diagnostic artifacts There was a mix of hard hammer and soft recovered to date are Early Archaic (early hammer percussion used in the blade Holocene). production, with variable skill level and execution (Table 2; Figure 8). Site The Technology of Production in the 40PT209 has also yielded numerous core Paleoindian-Early Archaic Transition edge flakes and crested blade fragments, and far more unifacial tools than any other Rock Creek Mortar Shelter is the first site on the UCP. Some evidence for shelter on the UCP where we have overshot flaking is represented by two encountered buried Early Archaic biface failures (one is an initial large deposits in good stratigraphic position. biface recovered in March 2013 that, due Early Archaic artifacts are very common in to a number of step fractures, was the region, but stratified sites are not. The reworked into an end scraper and used Early Archaic sequence extends from just for scraping hide). after the Younger Dryas (Greenbrier In short, there is blade technology at Dalton, Lost Lake and Kirk Corner the site along with bifacial reduction and Notched bifaces) through the end of the thinning. Numerous biface thinning flakes Early Archaic (bifurcate points such as St. have been recovered, and the vast Albans and LeCroy). All but one of our majority (about 80%) of the flaking debris early dates are associated with the Early and tools are made from locally available Archaic period, with both prismatic and Monteagle chert. Other identified raw triangular blades spatially associated with material types include varieties of Ft. these dates. Payne chert, St. Louis chert, and locally

67 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 8. Blade fragments recovered from Rock Creek Mortar Shelter. Table 2. Attributes for Selected Early Holocene Artifacts (40PT209).

available chalcedony. The entire range of and blades have come from a very lithic reduction is present in the early restricted area, about nine square meters, levels. Only one prismatic blade core and likely represent just a small sample of fragment (Figure 9) has been recovered materials in the shelter. from what appears to be an Early Archaic While questions remain about level (in Stratum 8). The authors stratigraphic associations, a shift in hypothesize that the function of some, if technology can be hypothesized. not most, blade cores may have shifted to Prismatic blade tool production (e.g., bifacial reduction at a certain point in the Paleolithic and Paleoindian) was a very chaîne opératoire. These early artifacts specific type of production involving very

68 Rock Creek Mortar Shelter

regular lateral margins (Figure 10). While prismatic blades tend to be curved in profile, these blade-like flakes are flat. The authors suggest this shift was intentional. Widely available raw materials may have allowed early inhabitants of the plateau the luxury of spending much less time preparing cores for stone tool manufacture in favor of more expedient methods for essentially the same end products.

Use-Wear Analysis of Selected Late Pleistocene/Early Archaic Materials

A functional use-wear study was undertaken as part of the multi- disciplinary investigation at 40PT209 that addresses the question of why people FIGURE 9. Blade core fragment recovered were using this shelter during the late from Rock Creek Mortar Shelter. Pleistocene. Traditional methods developed and refined in the 1980s by Keeley (1980) and Odell (1977) were employed to investigate the wear patterns. For this study, a Nikon Eclipse L150 metallurgical microscope was used to assess the damage and a Canon EOS Rebel T3i 18 mp dslr camera was used to document the findings. Each artifact was photographed on both surfaces so that the location of wear traces could be recorded directly on the image during analysis (Figure 11). The stone artifacts were lightly washed, with rubbing alcohol occasionally used to remove finger FIGURE 10. Long, flat blade-like flakes grease. Interpretations were made with recovered from Rock Creek Mortar Shelter. reference to an experimental collection precise core preparation and the (Hays 1998, Hays and Lucas 2001) that production of long, straight blades with has continued to develop over the last regular lateral edges. Such production is 20+ years with a wide variety of raw apparent at the site, but at a certain point materials. The experimental collection a production shift occurred. Long flakes represents a wide range of prehistoric were still selected, but the manufacturing activities, including projectile damage, process changed as cores were no longer butchering or meat processing, hide intricately prepared. The resulting flakes processing, bone and antler working, and were long and flat (bladelike) without the wood working.

69 Tennessee Archaeology 8(1-2) Summer 2016

the conservative side. To date, 109 pieces have been analyzed microscopically, with 18% (n=20) interpreted as used (Table 3). Represented activities included scraping wood (18%), processing hide, and cutting/scraping meat and hide (80%), and scraping bone (5%) which may be associated with meat processing. Three tools were used to scrape wood (a biface, a blade fragment with graver spurs and a Hardaway Side-Notched point), 16 were used to cut and scrape meat/hide in the early stages of processing, and a single scraper was used on bone. The represented activities are similar to those from the Nuckolls site (40HS60) on the lower Tennessee River (Ellerbusch 2004). Given the distance to the raw material sources (20 to 30 km), it might be FIGURE 11. Example of recorded wear - expected that a larger number of PP133, BCL14-011 - End Scraper on Blade retouched tools would be used. Also, in Use: Hide Scraping. consideration of the range of activities identified at Rock Creek Mortar Shelter, Table 3. Microwear Summary for Rock Creek the authors believe the lower use rate is a Mortar Shelter Late Pleistocene/Early reflection of post-depositional alterations Holocene Artifacts. along with the decision for conservative analysis interpretations.

Discussion

When considering the early Holocene The sediments in Rock Creek Mortar assemblage at site 40PT209, a major shelter are very dry, lending to good problem is that we do not yet know how to faunal and botanical preservation. characterize late Pleistocene/early However, there was microscopic evidence Holocene assemblages on the UCP, as of post- depositional alteration to the just a very small portion of Rock Creek surface of some stone tools that Mortar Shelter has been tested. The mix presented as an overall sheen. This of prismatic blade production with a more sheen was probably caused by movement expedient blade-like flake production is in these sandy deposits that so nicely puzzling. Further, select artifacts appear preserved bone. Every effort was taken to Paleoindian in character but were distinguish between polish resulting from recovered from an Early Archaic context. use and the naturally occurring sheen. One such artifact, a double side scraper With this phenomenon in mind, the made on a blade or flake fragment (Figure interpretations in this study lean toward 12, top row, right), would “fit” in the Clovis

70 Rock Creek Mortar Shelter

(2008) in their thorough survey of the Paleoindian period in Kentucky. In any case, regional comparisons are going to be difficult without comparable chronologies, and comparable chronologies may be tenuous without like assemblages. Stanfield Worley Bluff Shelter is a massive sandstone rock shelter with some 800m2 of potential living space located a little more than 10 km from the central Tennessee River Valley (Walthall 1998:230). DeJarnette and colleagues (1962) recorded a significant Dalton component in Zone D of the shelter deposits that they dated to between 10,000 and 9,000 years ago. More recently, Hollenbach (2009:101) obtained a calibrated AMS radiocarbon determination from Zone D that straddles the Pleistocene/Holocene boundary like Piece Plot 166 from Rock Creek Mortar Shelter. Preservation at Stanfield Worley FIGURE 12. Unifacial artifacts recovered from was excellent with numerous bone tools Rock Creek Mortar Shelter. and fauna recovered. Similar to Rock assemblage at the Carson-Conn-Short Creek Mortar Shelter, side scrapers on (40BN190) site (John B. Broster, personal blades, unifacial scrapers, and gravers communication, 2013). Clearly the site were present (Figure 12). There were also does not have a Clovis component myriad Dalton bifaces recovered (Walthall defined at this time, but the lithic 1998:230-231). However, thus far the assemblage is nonetheless interesting Rock Creek biface assemblage is because of such artifacts as gravers, comparatively sparse. gravers on end scrapers, and side We seem to have an Early Archaic scrapers on blades that are more typically assemblage at Rock Creek associated associated with Paleoindian lithic with blade (and biface) technology. The assemblages in the Midsouth (Broster et timing, if not the technology, is consistent al. 2006). with Walthall’s (1998) contention that In a recent synthesis, Anderson and Dalton peoples were the first to Sassaman (2012) place Early Archaic systematically use upland rock shelters. cultural sequences in the Southeast This also seems consistent with Lane and between 11,500 and 8,900 years ago. Anderson’s (2001) idea that the interior This seems to be at odds with some highlands (e. g., the Cumberland Plateau) regional chronologies that have middle lie at the end of early migration routes. In and late Paleoindian sequences any case, there are numerous recovered overlapping, e.g., the loose chronology artifacts that are Paleoindian in character proposed by Maggard and Stackelback if not in age. Bradbury and Carr (2012)

71 Tennessee Archaeology 8(1-2) Summer 2016

suggest that blade technology was not Job Site shelters are located on the East necessarily part and parcel of the Fork Obey River. Paleoindian toolkit, though they do point Franklin (2002) recorded early, middle, out that blade manufacture appears more and late Paleoindian sites (n=7) in common in Southeastern Paleoindian previous archaeological surveys of the assemblages. Bradbury and Carr (2012) UCP. Two sites may be classified as early further suggest that intentional (prismatic) Paleoindian through the recovery of blade manufacture was not part of the Clovis bifaces. No Cumberland bifaces Early Archaic tool kit at all. While Early were recorded during the survey. Five Archaic peoples made some blades and sites are late Paleoindian represented blade-like flakes, true blade (or prismatic primarily by Beaver Lake and Quad blade) technology should date earlier. And bifaces. One site contained a Dalton yet, at Rock Creek Mortar Shelter, there biface, and a local collector possesses a are blades in an early Holocene spectacular, very large Dalton biface from chronological context, so perhaps the area. To date, Rock Creek Mortar Bradbury and Carr’s contention should not Shelter is the only site of more than 600 be considered hard and fast. recorded through surveys that possesses Rock Creek Mortar Shelter is not the intact and stratified deposits. only late Pleistocene/early Holocene site we have recorded on the UCP. The Early Summary Times Rock Shelter yielded a late Paleoindian Quad biface with a heavily Rock Creek Mortar Shelter is a high ground base along with several wedge elevation rock shelter far removed from pieces (pièces esquillées) and an Early major streams and high quality, large Archaic MacCorkle Stemmed biface. package size raw material sources. In Interestingly, all of these items were fact, the closest possible raw material demonstrated to have been used for wood sources are at least 10 km away (and 200 working. The site was pristine, but the m below in elevation), and it is not clear deposits were largely surficial and thus these sources were exposed 11,000 not well dated. Further, no obvious blades years ago. Raw material surveys have not were recovered. Technological and been exhaustive due to the rugged terrain functional studies at Early Times Rock and vegetation cover, but the more likely Shelter indicate small groups of hunter- sources occur some 30 km to the south in gatherers used the shelter as a temporary the East Fork Obey gorge (Figure 13). special purpose locale where a limited This statement runs counter to ideas that number of activities took place. Locally prismatic blade production in the procured nodules of Monteagle chert were Tennessee region was anchored to more reduced at the site with tools occasionally localized outcrops of high quality cherts produced and resharpened. Also, some (Broster et al. 2013:304; Ellerbusch wood working took place and hide and 2004:36). Places such as Rock Creek meat processing of animals occurred on a Mortar Shelter may lie at the end of early small scale (Dye et al. 2010). migration routes, but it is unclear how Very near to Early Times Rock Shelter early peoples accessed the site since it is is the Job Site Rock House site where a nearly 30 km from the mouth of Rock Beaver Lake biface was recovered from a Creek at the Big South Fork of the disturbed context. The Early Times and Cumberland River in southeastern

72 Rock Creek Mortar Shelter

northwestwardly to the head of Wolf River [Pogue Creek] along the mountaintop to Kentucky." (Hogue 1933:2). A generally continuous record of human occupation at Rock Creek Mortar Shelter has been recorded from at least the end of the Pleistocene around 11,500 years ago to about AD 1000. The late Pleistocene and early Holocene deposits located 1.25 – 2 meters below surface have yielded many blades, unifacial side and end scrapers, gravers on scrapers, and a few bifaces from a restricted area under the drip line of the shelter. Broster et al. (2006:120) suggest these types of artifacts from the Widemeier site (40DV9) on the Cumberland River in the Central Basin were probably associated with Clovis and Cumberland components, though their excavation areas were often mixed. Dates and general stratigraphy at Rock FIGURE 13. Raw material sources in the vicinity of Rock Creek Mortar Shelter suggest Creek Mortar Shelter. these artifacts are associated Kentucky to the shelter near the with the Early Archaic period. However, headwaters of Rock Creek. If the better there is one AMS date of 12,554 cal BP in raw material sources were along the East (mixed) deposits that possess blade Fork Obey River (where the Early Times cores, prismatic blades, and other and Job Site rock shelters are also artifacts sometimes considered to be located), then it may be that early representative of early (Clovis) pioneering populations were moving up Paleoindian period assemblages in the Obey and across the ridge tops Tennessee (Broster et al. 2013:299; toward Rock Creek. State Route 154, Collins and Hemmings 2005). which runs near the site, follows the best Clearly there is much work to do to passable ridge tops up into Kentucky. sort out the stratigraphic, chronological, Local inhabitants indicate this route was and technological relationships at also an old Indian trail: "Another trail left 40PT209. For now, we might hypothesize the East and West Trail about three miles a transitional assemblage, or set of East of Jamestown and went assemblages. For example, the gravers

73 Tennessee Archaeology 8(1-2) Summer 2016

hunting camp occupied by residentially mobile foragers. Given the large size of the shelter, we hypothesize that this is consistent with Walthall’s (1998) ideas of Dalton rock shelter use. Site 40PT209 would fall in Walthall’s (1998) Group I shelters with an estimated 150m2 of living space. Early Holocene family groups likely used Rock Creek Mortar Shelter as a temporary residence during the fall and perhaps winter months. This article has briefly introduced the archaeology of Rock Creek Mortar Shelter. The authors are excited about additional work at a site with a continuous cultural sequence spanning more than 11,000 years. Of note is that one of the crew crawled back under the lower back ledge to a distance of about 20 feet (Figure 14). Given the depth of the cultural deposits already evaluated, the shelter likely extends much farther back and deeper during the Late Pleistocene FIGURE 14. Back ledge at Rock Creek Mortar and early Holocene. Shelter above buried deposits. and gravers on end scrapers suggest the Acknowledgements. Discovery, initial testing, and manufacture of spearshafts and foreshafts the 2014/2015 winter field season of Rock Creek Mortar Shelter was funded by three Tennessee (e. g., Broster et al. 2006:126). Historical Commission survey and planning grants. There also seems to be a mix of skill Funding of the 2013/2014 winter field season was level and execution for blade funded by National Geographic Society/Waitt manufacture. A few of the well-made Institute Grant No. W302-13. The winter blades would be at home in European 2013/2014 testing was conducted under Tennessee Division of Archaeology (TDOA) Late and Epi-Paleolithic assemblages, Permit No. 000796. The Tennessee Division of while other blades are poorly made. This Archaeology also generously funded two AMS may suggest a family group on site where radiocarbon dates. The folks at the Division (Mike older and more skilled knappers taught Moore, Mark Norton, John Broster, Suzanne younger novices to make blades. Also, Hoyal, and Aaron Deter Wolf) have been very supportive of our work at Pickett, and we are the earliest inhabitants of the UCP may thankful for their interest. We are especially have been coping with the constraints of grateful to Alan Wasik, Pickett State Park using the small rounded local cobbles of Manager, for his long term support of and Monteagle chert for blade production (as commitment to archaeological work at Pickett. opposed to large tabular cherts Brock Hill, Deputy Commissioner, Tennessee Department of Environment and Conservation has encountered in the lower Tennessee generously supported our work at Pickett and River drainage). Pogue Creek Canyon State Natural Area, The Rock Creek Mortar Shelter is particularly with arrangements for lodging. John currently interpreted as a short-term Froeschauer, Interpretive Specialist, Tennessee State Parks, has not only been supportive of the

74 Rock Creek Mortar Shelter

work but has also been a great friend and Tuscaloosa. colleague in the field. In general, we greatly appreciate of our relationship with Tennessee State Parks and their commitment to Broster, J. B., M. R. Norton, B. Hulan, and archaeological resource management. The MOUs E. Durham between Tennessee State Parks and ETSU 2006 A Preliminary Analysis of Clovis Archaeology are a perfect example of how through Early Archaic Components mutually beneficial partnerships in resource of the Widemeier Site (40DV9), management can be. Jay Franklin would like to thank David Todd, Assistant State Forester, Davidson County, Tennessee. Tennessee State Forests, for his support of and Tennessee Archaeology 2(2):120- interest in the project. Jay Franklin also thanks 127. Vice Provost Bill Duncan and the folks at ETSU Research and Sponsored Programs for their Broster, J. B., M. R. Norton, D. S. Miller, support and administration of the grant contracts for this project. Thanks also go to Joe Smith, J. W. Tune, and J. D. Baker ETSU University Relations, for his help in 2013 Tennessee’s Paleoindian Record. publicizing the work. We appreciate the interest In The Eastern Fluted Point and favorable reviews for the project from David Tradition, edited by J. A. M. Anderson and Kandi Hollenbach. Jacob Wall and Gingerich, pp. 299- 314. University Christina Bolte served as crew chiefs for the project. Andrew Hyder served as a graduate of Utah Press, Salt Lake City. assistant. Cayla Cannon conducted the ceramic analysis, and Sierra Bow assisted with the ceramic Collins, Michael B. and Andrew C. analysis. Jay Franklin would like to thank the many Hemmings ETSU students who have worked on the project to 2005 Lesser-known Clovis Diagnostic this point, and he greatly appreciates the support from ETSU in allowing us to conduct Artifacts I: The Bifaces. La Tierra archaeological field schools in the dead of winter. 32(2):9-20. Finally, we would like to acknowledge John Broster, who enthusiastically advised us DeJarnette, D. L., E. B. Kurjack, and J. W. concerning the character of the early Holocene Cambron assemblage from the site. 1962 Stanfield-Worley Bluff Shelter

Excavations. Journal of Alabama References Archaeology 8:1–111. Anderson, David G. and Kenneth E. Sassaman Des Jean, T. and J. L. Benthall 2012 Recent Developments in 1994 A Lithic Based Prehistoric Cultural Southeastern Archaeology. Society Chronology of the Upper for American Archaeology. The Cumberland Plateau. Tennessee SAA Press, Washington, D. C. Anthropologist 19(2):115-147.

Bradbury, Andrew P. and Phillip J. Carr Dye, Andrew D., Jay D. Franklin, and 2012 Evaluating Early Archaic Blade and Maureen A. Hays Bipolar Technologies. In 2010 Lithic Technology and Site Contemporary Lithic Analysis in the Function at Early Times Rock Southeast: Problems, Solutions, Shelter, Upper Cumberland and Interpretation, edited by Phillip Plateau, Tennessee. Paper J. Carr, Andrew P. Bradbury, and presented at the 67th Annual Sarah E. Price, pp. 79-95. Meeting of the Southeastern University of Alabama Press, Archaeological Conference,

75 Tennessee Archaeology 8(1-2) Summer 2016

Lexington, Kentucky. Questioning the Answers: Resolving Fundamental Problems Ellerbusch, Elijah C. of the Early Upper Paleolithic, 2004 Paleoamerican Prismatic Blade edited by Maureen Hays and Paul Economy from the Nuckolls Site Thacker, pp. 109-116. B.A.R (40HS60), Lower Tennessee River International Series 1005. British Valley, Tennessee. Current Archaeological Reports, Oxford. Research in the Pleistocene 21:35- 38. Hogue, A. R. 1933 One Hundred Years in the Franklin, Jay D. Cumberland Mountains: Along the 1999 The Rime of the Ancient Miners. Continental Line. The Standard MA thesis, Department of Printing Co., McMinnville, Anthropology, University of Tennessee. Tennessee, Knoxville. 2001 Excavating and Analyzing Hollenbach, K. D. Prehistoric Lithic Quarries: An 2009 Foraging in the Tennessee River Example from 3rd Unnamed Cave, Valley 12,500 to 8,000 Years Ago. Tennessee. Midcontinental Journal University of Alabama Press, of Archaeology 26(2), Cave Tuscaloosa. Archaeology in the Eastern Woodlands, pp. 199-217. Homsey-Messer, Lara 2002 The Prehistory of Fentress County, 2015 Revisiting the Role of Cave and Tennessee: An Archaeological Rockshelters in the Hunter- Survey. PhD dissertation, Gatherer Taskscape of the Archaic Department of Anthropology, Midsouth. American Antiquity University of Tennessee, Knoxville. 80(2):332-352. University Microfilms, Ann Arbor. 2006 Prehistoric Culture Chronology on Keeley, Lawrence the Upper Cumberland Plateau of 1980 Experimental Determination of Tennessee. Paper Presented at Stone Tool Use: A Microwear the 63rd Annual Southeastern Analysis. University of Chicago Archaeological Conference, Little Press, Chicago. Rock, Arkansas. Lane, L. and D. G. Anderson Hays, Maureen 2001 Paleoindian Occupations of the 1998 A Functional Analysis of the Southern Appalachians: A View Magdalenian Assemblage from from the Cumberland Plateau of Grotte XVI Dordogne, France. PhD Kentucky and Tennessee. In dissertation, Department of Archaeology of the Appalachian Anthropology, University of Highlands, edited by L. P. Sullivan Tennessee, Knoxville, Tennessee. and S. C. Prezzano, pp. 88-102. The University of Tennessee Hays, Maureen, and Géraldine Lucas Press, Knoxville. 2001 Experimental Investigations of Aurignacian Dufour Bladelets. In

76 Rock Creek Mortar Shelter

Langston, Lucinda M. Phillips, Jonathan D., Sarah McCormack, 2013 Site Location Modeling and Jidan Duan, Joseph P. Russo, Prehistoric Rock Shelter Selection Anne M. Schumacher, Ganesh M. on the Upper Cumberland Plateau Tripathi, Ruth B. Brockman, Adam of Tennessee. MA thesis, B. Mays, and Sruti Pulugurtha Department of Geosciences, East 2010 Origin and Interpretation of Tennessee State University, Knickpoints in the Big South Fork Johnson City. Electronic Theses River Basin, Kentucky-Tennessee. and Dissertations. Paper 1157. Geomorphology 114:188-198. http://dc.etsu.edu/etd/1157. Accessed 21 May 2016. Walthall, John A. 1998 Rockshelters and Hunter-Gatherer Langston, Lucinda and Jay D. Franklin Adaptation to the 2011 Archaeological Survey of Pogue Pleistocene/Holocene Transition. Creek State Natural Area: A GIS American Antiquity 63(2):223-238. Perspective. Paper Presented at the 23rd Annual Current Research Jay Franklin in Tennessee Archaeology East Tennessee State University

Meeting, Nashville. Maureen Hays College of Charleston Langston, Lucinda M., Jeffrey W. Navel, and Jay D. Franklin Frédéric Surmely 2012 Archaeological Survey in Pickett Ministère de la Culture DRAC Auvergne/SRA

State Forest, 2011-12. Paper Ilaria Patania Presented at the 24th Annual Boston University Current Research in Tennessee Archaeology Meeting, Nashville. Lucinda Langston U.S. Bureau of Land Management

Maggard, Greg and Kary Stackelback Travis Bow 2008 Paleoindian Period. In The Tennessee State Parks Archaeology of Kentucky: An Update, Volume 1, State Historic Preservation Comprehensive Plan Report No. 3, edited by David Pollack, pp. 109-192. Kentucky Heritage Council, Frankfort.

Odell, George 1977 The Application of Micro-Wear Analysis to the Lithic Component of an Entire Prehistoric Settlement: Methods, Problems, and Functional Reconstructions. PhD dissertation, Department of Anthropology, Harvard University, Cambridge, Massachusetts.

77 COLONIZATION AFTER CLOVIS: USING THE IDEAL FREE DISTRIBUTION TO INTERPRET THE DISTRIBUTION OF LATE PLEISTOCENE AND EARLY HOLOCENE ARCHAEOLOGICAL SITES IN THE DUCK RIVER VALLEY, TENNESSEE

D. Shane Miller and Stephen B. Carmody

While the timing of the initial colonization of North America is still hotly debated, the appearance of the Clovis culture likely represents an early, widespread colonization episode at the end of the Pleistocene. Here, we use the Ideal Free Distribution from Behavioral Ecology to interpret variability in the spatial distribution of previously recorded archaeological sites in the Duck River Valley in Middle Tennessee from the appearance of Clovis sites in the terminal Pleistocene though the Early Holocene (~13,250 – 8,880 cal yr BP). We hypothesized that the distribution of Clovis sites would be skewed towards the confluence of the Duck and Tennessee Rivers, and then subsequent populations would spread to higher elevations over the course of the Younger Dryas and Early Holocene as boreal forests were replaced by mixed hardwood, deciduous forests. After correcting the sample of archaeological sites to account for survey and taphonomic biases, we found that sites dating to the latter part of the Younger Dryas and the Early Holocene become more frequent at higher elevations. However, contrary to the predictions of our model, site frequencies become less frequent during the Early Holocene at the confluence of the Tennessee and Duck Rivers. Our results are consistent with other studies that have proposed that the Cumberland Plateau and the Appalachian Highlands were not intensively occupied until well after the disappearance of the Clovis culture.

In this paper, we examine Late 1999; Smith 1995, 1996), and other major Pleistocene and Early Holocene changes that occurred in eastern North landscape use in the Duck River drainage America during the mid-Holocene using a model from behavioral ecology, (Anderson et al. 2007). Additionally, the Ideal Free Distribution (Fretwell and archaeologists have conducted multiple Lucas 1970). Understanding how hunter- large scale surveys in the Duck River gatherers utilized the Duck River drainage drainage, making it one of the most during the Pleistocene-Holocene intensively studied areas of the Mid-South transition is particularly important because and an ideal location to track variation in of its location in the heart of the Midsouth, landscape use (Miller 2014). a region that likely played a vital role in Using a geospatial database of not only the colonization of the Americas, archaeological sites derived from the state but also the first steps in a trajectory that site file records at the Tennessee Division eventually led to increasing sedentism of Archaeology and the Tennessee Fluted and resource intensification (Bissett 2014; Point Survey (e.g., Broster et al. 2013; Brown and Vierra 1983), plant Miller 2014), we found that archaeological domestication (Chapman and Shea 1981; sites were initially concentrated in the Crites 1985, 1987, 1993; Smith 1987, confluence of the Duck River with the 1992, 2001, 2007, 2011), emerging trade Tennessee River. However, over the networks (Jefferies 1995, 1996, 1997; course of the Younger Dryas and into the McNutt 2008), inter-group violence (Milner Early Holocene, the distribution of

78 Colonization after Clovis archaeological sites extends to the collectors and avocational archaeologists Eastern Highland Rim and the recording the location and attributes of Cumberland Plateau. We argue that this Paleoindian period projectile points dating pattern is consistent with broader trends back to the mid-1940s with Ben McCary’s in landscape use (e.g., Lane and (1947) “Folsomoid” point survey in Anderson 2003; Maggard and Virginia. Mason (1962), Brennan (1982), Stackelbeck 2008; Tankersley 1990), and and later Anderson (1990) compiled this reflects populations expanding into higher information and used their results to elevations as temperate forests replaced create distribution maps that were the boreal forests as climate became warmer basis for creating models of landscape and wetter during the Early Holocene. use and colonization processes. Because of their efforts, we now know that there is Paleoindian and Early Archaic a rich Pleistocene and Early Holocene Landscape Use in the Southeastern archaeological record in the southeastern United States United States, and contemporary researchers have at their disposal a When people discuss the colonization dataset that allows them to analyze of the southeastern United States, it is Paleoindian landscape use at multiple common practice to begin with a temporal and spatial scales. discussion of the Clovis culture, and the The ability to evaluate changes in merits and shortcomings of a handful of landscape use is particularly important in pre-Clovis sites like Cactus Hill, Paleoindian archaeology, where major Meadowcroft Rockshelter, Page-Ladson, research themes include the timing, rate, and Topper (Anderson et al. 2015; and nature of the spread of peoples into Goodyear 2005). Aside from these sites, North America at the end of the there are only a few sites with buried, Pleistocene (Meltzer 2009:1-22). In terminal Pleistocene archaeological particular, Martin (1973, 2006) and deposits, and an even smaller percentage Haynes (1964, 2005) argued that Clovis of these that have produced widely were the initial colonists that spread accepted radiocarbon dates. rapidly across North America with Consequently, archaeologists in the relatively high reproductive rates while southeastern United States have to rely hunting megafauna species to extinction. heavily on well-dated sites from outside This model is based primarily on the the region to construct chronologies, as widespread distribution of Clovis the climate and geomorphology of the technology across North America and the Southeast is not conducive to site burial recovery of multiple kill sites on the Great and organic preservation (Dunnell 1990; Plains and in the southwestern United Goodyear 1999; Miller and Gingerich States. Kelly and Todd (1988) updated 2014). the “Clovis First/Overkill” model to argue However, the southeastern United that the first colonists of the Americas States is unique in that, while there are were “technology-oriented.” In other only a relatively few dated sites, there is words, Clovis groups coupled high an abundance of surface sites and residential mobility with a toolkit geared isolated finds (Anderson et al. 2010). towards hunting large animals to Moreover, professional archaeologists in overcome any environmental incongruities the region benefit from a long history of they encountered.

79 Tennessee Archaeology 8(1-2) Summer 2016

The archaeological record of eastern place-oriented colonization strategy would North America, and in particular the Mid- have the benefit of minimizing resource South, is often cited in critiques of “Clovis uncertainty and demographic risk. First / Overkill / Technology-Oriented” The distribution and frequency of models for the colonization of the Paleoindian projectile points in the Americas. Critiques generally reference southeastern United States have also the antiquity of sites in eastern North been used in arguments regarding America that pre-date Clovis (e.g., whether or not the onset of the Younger Goodyear 2005), the limited evidence for Dryas stadial had a discernable effect on predation on megafauna aside from people. Anderson and Faught (2000) Kimmswick in Missouri (Graham et al. used the distribution of bifaces to argue 1981), Coates-Hines in Tennessee that, at the onset of the Younger Dryas, (Deter-Wolf et al. 2011), and Big Bone the range of projectile point types Lick in Kentucky (Tankersley et al. 2009), diminished and regionally specific types and the broad diet breadth exhibited in the appear to take the place of Clovis. They Late Paleoindian components at Dust argued that this likely represents a Cave in Alabama (Hollenbach 2007, fracturing of information exchange 2009; Walker 1998, 2007). However, the networks that appear to coincide with primary appropriation of the eastern abrupt and sustained climate change. Paleoindian record to critique the “Clovis Similarly, Goodyear (2006) and Daniel First/Overkill/Technology-Oriented” model and Goodyear (2006) argue that a decline is the sheer density and differential in the frequency and distribution of distribution of archaeological sites and Redstone bifaces (presumably a post- artifacts. Gardener (1974, 1983) argued Clovis projectile point) likely indicates a that Clovis sites were clustered, or were decline in population at the onset of the tethered, to lithic raw material sources in Younger Dryas. However, Smallwood the Shenandoah Valley of Virginia. (2012) has identified subtle variation in Meltzer (1988) interpreted the frequent the production of Clovis bifaces in areas occurrences of isolated artifacts and that Anderson identified as staging areas, surface scatters in the southeastern leading her to conclude that emerging United States as indicative of a forager divisions in learning networks were well subsistence strategy (e.g., Binford 1980), underway prior to the onset of the as opposed to a collector strategy Younger Dryas. Similarly, Thulman (2009) displayed by archaeological sites at found that early Paleoindian period higher latitudes in eastern North America. bifaces in Florida were remarkably similar, Anderson (1990, 1995; Anderson and whereas subtle variation in the variability Gillam 2000) used the distribution of of Late Paleoindian types clustered Paleoindian period projectile points as the geographically, which he also argued was basis for his “staging area” model for the indicative of the fracturing of learning colonization of North America. He networks. Moreover, he also found that contends that groups settled in resource- the distribution of sites during the Younger rich valleys, and then expanded in a leap- Dryas was located near sinkholes, leading frog fashion into subsequent areas. him to conclude that access to freshwater Anderson’s “staging area” model dovetails may have been an issue during the with a demographic model put forth by Younger Dryas. Consequently, the effect Meltzer (2004), who contends that a of the onset of the Younger Dryas in the

80 Colonization after Clovis southeastern United States is still an open different tool types, he argued that Early question, but it is clear that the distribution Archaic mobility consisted of large, and frequency of archaeological sites and residential base camps buried in the isolated finds has, and will continue, to alluvial terraces, while smaller logistical play a role in resolving this issue. camps were situated on the periphery of More specific to the Mid-South, the the drainage. A similar study conducted in distribution of Dalton bifaces in Arkansas the Little Tennessee River Valley by Davis has been at the center of a debate as to (1990) showed that base camps were whether Late Paleoindian groups moved more often located along the first terrace within, or between, river drainages (Gillam of the Little Tennessee River while 1999; Morse 1971, 1973, 1975, 1977; logistical camps, when identified were Schiffer 1975a, 1975b). Walthall (1998) found on the back edge of the Little argued that by the end of the Pleistocene, Tennessee River valley or along its Dalton groups’ use of rockshelters and the tributaries (Davis 1990:210). A shift cemetery site at Sloan indicate that towards the use of the uplands occurred groups were clearly “place-oriented” and during the late Early Archaic period, used the same locations repeatedly. identifies as the Upper Kirk period. The Hollenbach (2007, 2009) used variability trend quickly reverses by the end of the between paleobotanical assemblages to Early Archaic period where base camp argue that terminal Pleistocene and Early sites are again primarily found along the Holocene groups in northern Alabama Little Tennessee River valley (Davis moved around the landscape in a well- 1990:210). Kerr and Bradbury (1998) developed, and well-planned, seasonal argued that in the Lower Tennessee River cycle to take advantage of predictable Valley, Early Paleoindian assemblages fluctuations in resource availability. were small and dispersed, while Terminal Maggard and Stackelbeck (2008) contend Paleoindian and Early Archaic sites were that the rarity of early sites on the greater in number and reflected a wider Cumberland Plateau in Kentucky likely use of the valley. Anderson and Hanson indicates that it was not colonized until (1988) used raw material variation and the latter parts of the Younger Dryas and into ratio of expedient to curated tools to argue the Early Holocene, a pattern that Lane that bands of Early Archaic hunter- and Anderson (2001) have argued gatherers aggregated at the “Fall Line” extends for the entirety of the Cumberland during the fall and winter and dispersed to Plateau and the Appalachian Highlands. Piedmont and Coastal Plain during the Other attempts to discern patterns of spring and summer to take advantage of early hunter-gatherer landscape use in seasonally abundant resources. Daniel the southeastern United States utilized (2001) argued, similar to Gardener’s archaeological data from large federally (1974, 1983) model for the Shenandoah funded projects. However, most had only Valley, that rather than being oriented small samples of Pleistocene-aged sites along major rivers, Early Archaic or artifacts, and instead focused on Early settlement in the southeastern United Archaic assemblages. For example, States was tethered to lithic raw material Kimball (1996) utilized assemblages from sources. deeply stratified sites and isolated finds The heavy reliance on the analysis of from the Little Tennessee River Valley in site and artifact distributions has been East Tennessee. By using frequencies of ongoing theme for the study of the Late

81 Tennessee Archaeology 8(1-2) Summer 2016

Pleistocene and Early Holocene substantially more research than others, archaeological record in eastern North which leads to a form of survey and America. This is likely due to a research bias in the samples of convergence of factors that include: (1) Paleoindian and Archaic site distributions the influence of processual approaches (Anderson 1996; Buchanan 2003; that emphasize the analysis of the Prasciunas 2011; Shott 2002, 2005). organization of technology across time Another issue noted by Anderson (1996) and space in relation to environmental is the variation in the size of sites (isolated parameters (Binford 1979, 1980; Nelson finds vs. large, dense sites), and that state 1991); (2) a concerted effort to connect site files usually only record temporal data the early record of the southeastern by broad time periods. One way to resolve United States to larger questions this issue would be to convert a site- surrounding the colonization of the based approach into a non-site survey Americas (Anderson et al. 2015); (3) an (e.g., Dunnell and Dancey 1983; Thomas interest in the relationship between 1975). For example, Cabak et al. (1998) climate and culture change, especially in analyzed the site records for the regards to the impact of the Younger Savannah River site to make inferences Dryas stadial; and (4) an effort to make about shifts in landscape use over the the most out of a regional record has few duration of the Holocene. Another buried sites, but lots of surface scatters potential analytical avenue would be to and isolated finds (Dunnell 1990; use Surovell’s (2009) proxies for Goodyear 1999; Miller and Gingerich measuring occupation intensity for a 2013). By focusing on the analysis of site sample of assemblages. However, and artifact distributions, archaeologists attempting to replicate Cabak et al.’s studying the Pleistocene and Early (1998) and Surovell’s (2009) analyses Holocene record in the southeastern would be difficult because different United States have historically made the excavation protocols and curation criteria most out of the proverbial cards that they would severely limit the sample of sites have been dealt. available for analysis in the Duck River Valley. Using Site Distributions to Reconstruct Finally, attempts to use the distribution Demography and Landscape Use and frequency of sites to reconstruct demography suffer from a lack of a While using the frequency and developed interpretive framework. For distribution of sites and artifacts to assess example, Meltzer (1988) interpreted the demography and landscape use is high frequency of recorded Early common practice in archaeology, there Paleoindian sites and their wide are still several issues that must be distribution across the southeastern overcome with this approach. First, like United States as evidence that groups the frequency of 14C dates, archaeological were smaller and more residentially sites are also subject to a variety of mobile when compared to Paleoindian taphonomic biases that make finding and groups at higher latitudes. Anderson recording older archaeological sites more (1996) interpreted the increase in the difficult (e.g., Dunnell 1990; Kelly et al. frequency and wider distribution of sites in 2013; Surovell et al. 2009). Moreover, the Late Archaic across the southeast as some areas have been subjected to evidence for an increase in population

82 Colonization after Clovis

compared to earlier time periods. Does the increase in frequency and a wider distribution mean more people, or fewer people moving around more often? Clearly, there is an equifinality issue here. In this article, we argue that a formal model from behavioral ecology, the Ideal Free Distribution (Fretwell and Lucas 1970), can be appropriated as an interpretive framework for analyzing variation in site frequencies and distributions to understand fluctuations in landscape use, and by extension demographic trends. We apply this approach to a database of archaeological sites in the Duck River drainage to trace variation in landscape use over the Late Pleistocene and Early Holocene epochs.

The Ideal Free Distribution

The Ideal Free Distribution (IFD) was first used to model habitat selection by birds (Fretwell and Lucas 1970; Figure 1). Its most basic assumption is that individuals will select habitats to maximize fitness, and that the suitability of the habitat and population density will FIGURE 1. The Ideal Free Distribution. The influence an individual’s decision to either upper panel shows suitability curves on a stay in a habitat or move to a location with normalized scale of 0-1 for three habitats (A, greater net fitness benefits. Consequently, B, and C) as a function of population density if the IFD holds true, the distribution of in the habitat. The habitats are ranked in individuals should reflect the relative alphabetical order by the suitability experienced by the initial occupant, and in all suitability of habitats. The model assumes cases suitability declines with population that the first individuals moving into a new growth. The lowest ranked habitat (C) also area will occupy the most favorable experiences the lowest rate of declining habitat. As the quality of that habitat or suitability. The lower panel shows how patch declines due to increasing population growth will be allocated among population, resource depletion, and habitats given these suitabilities (Fretwell and competition, individuals will move to a Lucas 1970; Winterhalder et al. 2010:473) neighboring habitat that was initially less have all of the available information to suitable but is now superior to the original make a decision on whether to move or habitat. The IFD model shares certain stay, and that all individuals are free to characteristics with the Marginal Value leave or enter a new habitat. However, Theorem (e.g., Charnov 1976; Kelly unlike the Marginal Value Theorem, the 1995:90-97): it assumes that individuals Ideal Free Distribution incorporates

83 Tennessee Archaeology 8(1-2) Summer 2016

archaeological sites is determining the suitability of each habitat. Complicating matters further, this study uses assemblages that span several millennia where substantial environmental changes occurred in eastern North America (e.g., Delcourt and Delcourt 1981, 1983; Viau et al. 2006; Williams et al. 2004) (Figure 2). Delcourt and Delcourt (1981, 1983) argued that the boreal forests that covered eastern North America during the Last Glacial Maximum (LGM) began moving north and more temperate species expanded to cover large swaths of area by the Mid-Holocene as a response to increasing temperature and moisture. Anderson Pond (Delcourt 1979) in Tennessee FIGURE 2. Temperature reconstruction based on the and Jackson Pond in Kentucky Greenland Ice Sheet Project (GISP2) (Alley 2000) and (Wilkins et al. 1991) are the two pollen records (Viau et al. 2006) with the sequence of primary pollen cores that were temporally diagnostic projectile points (from Anderson and used to reconstruct the Sassaman 2012). transition from boreal to expectations relating to population deciduous forests in the Mid-South. A density, or landscape in-filling, which subsequent reanalysis of these cores by makes it useful as an interpretive Liu et al. (2013) placed the transition from framework for assessing demography and boreal-to-deciduous forests at 15,900 cal landscape use. The applicability of using yr BP at Anderson Pond and 15,400 cal yr the IFD in conjunction with archaeological BP at Jackson Pond. This transition does data has been demonstrated with recent not occur at Silver Lake in Ohio and studies of the transition to agriculture in Appleman Lake in Michigan until Spain (McClure et al. 2006), the spread of approximately 2,000 years later (Gill et al. agriculture into Europe (Shennan and 2009), which Liu et al. (2013) cite as Edinborough 2007), and the colonization support for a time-transgressive shift from of Oceania (Kennett al. 2006) and the in forest composition at the end of the Channel Islands in California (Kennett and Pleistocene from south to north. Winterhalder 2008; Winterhalder et al. While a time-transgressive shift from 2010). boreal to deciduous forests occurred A critical component of using the IFD northward, it also likely occurred upward. to interpret the distribution of In other words, the trend of shifting forest

84 Colonization after Clovis composition is also reflected in the distribution of species relative to altitude. For example, Mills and Delcourt (1991) observed that several areas of the Blue Ridge Mountains show evidence of alpine tundra being replaced by boreal forests after 12,500 years ago (~14,761 cal BP). Consequently, the replacement of these boreal forests by FIGURE 3. Variation in elevation across the Duck River deciduous forests likely Drainage derived from the National Elevation Dataset (USGS 2014) (vertical exaggeration 10x). occurred well after disappearance of alpine tundra. al. 1991). Furthermore, at Anderson pond Moreover, the reconstructions by Delcourt the influx of oak pollen not only correlates and Delcourt (1981, 1983) and Williams et strongly with hickory pollen, but also al. (2004) show a delay in the northward positively correlates with 25 other species, movement of boreal forests in areas with which contrasts significantly with the higher altitude, especially the Appalachian coniferous species that dominated Highlands and the Cumberland Plateau. eastern North America during the LGM. This pattern likely indicates a “sky island” The spread of mixed hardwood forests effect where remnant boreal forests and would have had the combined effect of alpine tundra are present in the uplands increasing the amount of oak and hickory well after deciduous forests spread into available for humans, as well as deer lower elevations. However, with the populations whose preferred food dramatic increase in temperature at the resource are acorns (Munson 1986; end of the Younger Dryas and into the Gardener 1997). Consequently, the Early Holocene, the size of these boreal expected pattern for settlement in the sky islands would have diminished as Duck River drainage would be for groups deciduous forests climbed to higher to colonize the lowest elevations first, elevations. This pattern is relevant to an followed by an expansion to higher analysis of the Duck River Valley because elevations as boreal forests give way to there is considerable variation in elevation deciduous forests from the Late from the eastern Gulf Coastal Plain Pleistocene to Early Holocene. (~115m) to the Cumberland Plateau (~600m) (Figure 3). Study Sample The shift from boreal to temperate forests would have greatly altered the Works Progress Administration crews resources available for prehistoric groups. first surveyed the areas around the For example, the pollen cores from both confluence of the Duck and Tennessee Jackson Pond in Kentucky and Anderson Rivers in the 1930s as part of the Pond in Middle Tennessee show large Kentucky Lake project (Lewis and increases in oak pollen from the Late Kneberg 1959). This area has since been Pleistocene through Early Holocene resurveyed by both Cultural Resource (Delcourt 1979; Liu et al. 2013; Wilkins et Analysts in the early 1990s (Kerr and

85 Tennessee Archaeology 8(1-2) Summer 2016

Bradbury 1998) and most recently by the University of Tennessee Archaeological Research Lab (Angst 2012). The section of the river in Coffee and Bedford counties in the Eastern Highland Rim was surveyed as part of the Normandy Reservoir project in the early 1970s (Faulkner and McCullough 1973). The area that traverses the Nashville Basin in Maury and Marshall counties was surveyed as part of the Columbia Reservoir project (Evans 1972). Jolley (1980) conducted a survey of FIGURE 4. The distribution of recorded archaeological sites in Hickman and sites in the Duck River drainage. Humphreys counties to locate sites Tennessee also has one of the most between the extent of the Kentucky Lake active Paleoindian projectile point surveys and Columbia Reservoir surveys. In the in North America (Anderson et al. 2010; ensuing years, many additional sites were Broster and Norton 1996; Broster et al. reported as part of cultural resource 2013). This database, obtained in large management projects, making this one of part from private collections, contains the more comprehensively surveyed information on the projectile point type, drainages in the Mid-South. The Duck metric attributes, raw material type, and River also crosses multiple physiographic date of recovery. Miller (2014) integrated provinces, including the lower Tennessee these sources into a database that River Valley, the Western Highland Rim, contains information for 2,427 recorded Nashville Basin, Eastern Highland Rim, sites from the eight counties spanning the and Cumberland Plateau (e.g., Fenneman extent of the Duck River drainage, of 1938). Because of the unique geologic which he was able to acquire specific setting and variation in elevation in this coordinates for 2,211 of these sites (Table drainage, we used each physiographic 1; Figure 4). For this study, we use 421 section as a proxy for a “habitat” in an components from 1175 terminal Ideal Free Distribution analysis. Paleoindian and Archaic period sites There are two primary sources of data based on the reported presence of on the distributions of sites and artifacts temporally diagnostic projectile point for the study area. The first is the types. Tennessee state archaeological site files at the Tennessee Division of Archaeology. Methods This dataset contains information on site location, condition, cultural affiliation, Miller (2014) assessed potential artifacts recovered or observed, and sources of sampling bias in the where the collections are curated. In distribution of archaeological sites. These addition to site file information, the state of potential effects that were assessed

86 Colonization after Clovis

Table 1. Summary Data for the Site File Record Analysis for the Duck River Drainage.

1 A component is determined by the presence of absence of a temporally diagnostic artifact at a site as defined by the Tennessee Division of Archaeology. included a taphonomic bias, modern components, which provides a relative population/collector bias, landcover, measure of the distribution of components geology, and geomorphology. Based on over time for that physiographic section. this analysis, a temporal taphonomic bias This is essentially the same approach and an over-representation of sites in Anderson (1996) used in his analysis of areas that were surveyed as part of the the distribution of Archaic sites across the Kentucky Lake, Columbia, and Normandy southeastern United States. reservoir projects are the two main Alternatively, we divided each cell by sources of variation that appear to be the total number of sites for the same biasing the study sample. temporal component. Again, using the Consequently, it is necessary to adjust Clovis sample from the Coastal the study sample in two ways to winnow Plain/Western Valley physiographic out the effects of these biases in the section as an example, we divided the reporting of sites containing Paleoindian total number of Clovis components in the and Archaic components in the Duck physiographic section (43) by the total River drainage. As a means of number of Clovis components in the entire counteracting survey bias, we first divided study sample (53). This provides a each cell (i.e. the frequency of relative measure of the distribution of components by time period and components across space at a particular physiographic section) by the total time. Modifying the sample in this way number of Paleoindian and Archaic helps to account for the taphonomic bias components identified in each that affects the temporal distribution of physiographic section. For example, in the sites by isolating the study sample at each Coastal Plain/Western Valley time slice to analyze how each physiographic section, there are 43 Clovis component varies across space. Finally, components out of a total of 309 we used Pearson’s “Goodness of Fit” test Paleoindian and Archaic components for (Pearson 1900) with Cramer’s (1946) V to the entirety of the Duck River drainage. control for scalar effects to determine if For that cell, we simply divided the the distributions are statistically number of Clovis components by the total significant. number of Paleoindian and Archaic

87 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 5. Distribution of temporal components divided by the total Paleoindian and Archaic components from within the physiographic section (column).

Results illustrate the degree to which sites are distributed across the drainage. As a null Two trends are apparent after hypothesis, we used an even distribution adjusting the data to account for survey of sites across the drainage. To derive coverage bias (Figure 5). First, the this, we divided the total number of majority of Clovis, Cumberland/Barnes, components for each time slice by the Quad/Beaver Lake, and Dalton/Harpeth percentage of the total area of the county. River/Greenbrier components are found in In other words, if a physiographic section Coastal Plain/Western Valley encompasses 30% of the total area of the physiographic section associated with the drainage, it should also contain 30% of lower Tennessee River Valley. However, the components for a given a time slice if there is a slightly higher representation of they were perfectly dispersed across the the Quad/Beaver Lake and Dalton/ drainage. To meet the minimum sample Harpeth River/Greenbrier components size requirements for each cell, we across the drainage, and in particular in combined the physiographic sections into the Eastern Highland Rim in Coffee three categories (Coastal Plain/Western County. Then, with the Early Archaic Highland Rim, Nashville Basin, and Corner-Notched and Bifurcate Eastern Highland Rim/Cumberland components, the trend reverses with the Plateau). We also collapsed the Clovis, majority of the sites in these time periods Cumberland, Quad/Beaver Lake and located in the Nashville Basin, Eastern Dalton/Greenbrier components into a Highland Rim, and Cumberland Plateau. general Paleoindian group, and the Early The same two trends also appear after Archaic Corner-Notched and Bifurcate adjusting the data to account for temporal components into another. This analysis taphonomic bias (Figure 6). illustrates that the trends described above We used Pearson’s “Goodness of Fit” are statistically significant. The test with Cramer’s V as another means to Paleoindian group is over-represented in

88 Colonization after Clovis

FIGURE 6. Distribution temporal components by county divided by the total components from the same temporal group (row). the Coastal Plain/Western Highland Rim Additional studies by Franklin (2002) in physiographic section and under- Fentress County and Jolley (1979) in the represented in the Eastern Highland Calfkiller River Valley along the western Rim/Cumberland Plateau (X2=42.85; df=3; edge of the Cumberland Plateau also p<.001; V=.32). The inverse pattern was show substantially more sites with Early observed in the Early Archaic group, with Archaic components compared to the frequency of sites over-represented in Paleoindian period components. This the Eastern Highland Rim/Cumberland pattern also extends into Kentucky, where Plateau and under-represented in the Early Paleoindian sites are rarely found Coastal Plain/Western Highland Rim on the Cumberland Plateau compared to physiographic section (X2=85.48; df=3; adjacent physiographic sections at lower p<.001; V=.45). elevations (Maggard and Stackelbeck 2008). Nor is this pattern restricted to the Discussion Cumberland Plateau. Both Bass (1977) and Kimball (1996) reported a large After correcting for temporal number of Early Archaic sites from the taphonomic bias and differential survey Great Smoky Mountains and Tellico coverage, the Clovis and Cumberland Reservoir in east Tennessee, of which sites are primarily found at the confluence Kimball used 333 different assemblages with the Tennessee River, and then to model Early Holocene site function and during the Middle and Late Paleoindian landscape use. Considering that the they start to appear with greater Tellico Reservoir and Great Smoky frequency in the Eastern Highland Rim. Mountains have been extensively By the Early Archaic period, sites are surveyed, it is striking that only seven much more prevalent in the Eastern projectile points (a Clovis, a Redstone, Highland Rim and Cumberland Plateau. two Cumberlands, two Quads, and one Moreover, this pattern does not appear Dalton) from Monroe, Blount, and Sevier to be unique to the Duck River Valley. counties are included in the Tennessee

89 Tennessee Archaeology 8(1-2) Summer 2016

Fluted Point Survey (Broster et al. 2013). confluence of the Tennessee and Duck Lane and Anderson (2001) argue that this Rivers. While Kirk Point, a substantial pattern extends for the entirety of the Early Archaic site, has been recorded in Appalachian Highlands and the this area (McNutt et al. 2008), there are Cumberland Plateau. fewer reported Early Archaic components Lane and Anderson (2001) and in this area when compared to the sites Maggard and Stackelbeck (2008) both compared to the rest of the Duck River argued that due to topography the drainage. Furthermore, Miller (2014) Cumberland Plateau and Appalachian created a database of over 5,000 bifaces Highlands were at the periphery, of the from Benton and Humphreys counties. He initial colonization process for eastern found relatively few Early Archaic Corner- North America. In both of these studies, Notched and Bifurcate projectile points they argued that areas at higher compared to the frequency of the elevations were perhaps ignored initially preceding Late Paleoindian Dalton types, because of the energy costs associated and the subsequent terminal Early with exploring areas at higher elevations. Archaic Kirk Stemmed and Middle Archaic Similarly, we argue that the relative Eva I types. The Ideal Free Distribution absence of Paleoindian sites and would predict that if the suitability of this subsequent abundance of Early Archaic habitat remained constant relative to other sites at higher elevations meets the habitats in the drainage, then the relative expectations of the Ideal Free Distribution frequency of sites in this area should also (e.g., Fretewell and Lucas 1970). This remain constant. However, that is clearly model assumes that the first individuals not the case. moving into a new area will occupy the There could be several explanations most favorable habitat, and as more for the decrease in the frequency of Early people move into this habitat, crowding, Archaic sites at the confluence of the resource depletion, and competition will Tennessee and Duck Rivers in the Early result in people colonizing secondary, Holocene. First, there could be unsettled habitats. However, rather than geomorphological bias specific to the the energetic costs of moving to higher Tennessee River that caused the erosion elevations being the primary deterrent, we of levee deposits utilized by Early argue that the composition of forests, Holocene hunter-gatherers. However, for which also correlates with elevation, was this to be the case, this process would the likely culprit that prevented early have to have also left the deposits populations from initially establishing a containing Late Pleistocene and Middle permanent presence in highland areas in Holocene intact. Second, there could also the Late Pleistocene. In other words, it be a chronological issue, where in the was not until regional climate became Tennessee River Valley Early Archaic warmer and wetter during the Early projectile point types are not mutually Holocene that mixed hardwood forests exclusive and overlap considerably in moved into higher elevations to displace time. Given the vagaries with the Late remnant boreal forests, thereby increasing Pleistocene and Early Holocene the abundance and diversity of resources radiocarbon record in the southeastern available for hunting and gathering. United States, placing accurate and More puzzling, however, is the precise time brackets for the appearance relatively few early Holocene sites at the and disappearance of temporally

90 Colonization after Clovis diagnostic projectile point types is still with the Tennessee and Duck Rivers. difficult (e.g., Anderson et al. 2010; However, over the course of the Younger Goodyear 1999; Miller and Gingerich Dryas and into the Early Holocene, the 2013). Finally, the decrease in sites dating distribution of archaeological sites to the Early Holocene could be due to extends to the Eastern Highland Rim and variability in local environmental the Cumberland Plateau. We argue that conditions that are not reflected in broader this pattern is consistent with broader studies (Delcourt and Delcourt 1981, trends in landscape use (e.g., Lane and 1983; Liu et al. 2013; Williams et al. 2004) Anderson 2003; Maggard and or paleoenvironmental proxies from Stackelbeck 2008; Tankersley 1990), and adjacent physiographic provinces reflects populations expanding into higher (Brackenridge 1984; Delcourt 1979; elevations as temperate forests replaced Klippel and Parmalee 1982; Wilkins et al. boreal forests as climate became warmer 1991). Despite the extensive surveys and and wetter during the Early Holocene. subsequent research conducted on sites However, counter to the expectations of from the Lower Tennessee and Duck the Ideal Free Distribution, we also River Valleys, there is still much to learn observed that the relative frequency of regarding Late Pleistocene and Early Early Holocene sites decrease at the Holocene chronology, environments, and confluence of the Tennessee and Duck hunter-gatherer adaptations in this area, River valleys. We argue that this may be and is critical for understanding the origins the result of Early Holocene erosion of the of the Shell Mound Archaic in the Mid- levees adjacent to the Tennessee River, South (e.g., Bissett 2014). imprecision in the regional sequence of temporally diagnostic projectile points, or Conclusion local environmental variability that has gone undetected in regional proxy Analyses of the distribution of records. archaeological sites have played a prominent role in the study of the Late Acknowledgements. We would like to thank Pleistocene and the Early Holocene Suzanne Hoyal for her help with Tennessee State Files, and John Broster and Mark Norton for epochs in the southeastern United States. making the Tennessee Fluted Point Survey We argue that the reliance on analyzing publicly available. This research was partially site distributions stems from the abundant supported by a dissertation improvement grant and widespread record of surface sites awarded by the SRI Foundation. We would both and isolated finds in the region in contrast like to thank John Broster again for allowing us to pester him about archaeology for over a decade. to the poor record of sites with buried components with organic material References remains. In this paper, we utilized the Ideal Free Distribution from Behavioral Alley, Richard B. Ecology (e.g., Fretwell and Lucas 1970) to 2000 Ice-Core Evidence of Abrupt analyze variation in landscape use in the Climate Changes. Proceedings of Duck River drainage. After adjusting the the National Academy of Sciences data to account for variation caused by 94(4):1331-1334. survey bias and a temporal taphonomic bias, we found that archaeological sites Anderson, David G. are initially concentrated in the confluence 1990 Paleoindian Colonization of

91 Tennessee Archaeology 8(1-2) Summer 2016

Eastern North America: A View River Valley. American Antiquity from the Southeastern United 53:262-286. States. In Early Paleoindian Economies of Eastern North Anderson, David G., D. Shane Miller, America, edited by K. Tankersley Stephen J. Yerka, J. Christopher and B. Isaac, pp. 163–216. Gillam, Erik N. Johanson, Derek T. Research in Economic Anderson, Albert C. Goodyear and Anthropology Supplement 5. JAI Ashley M. Smallwood Press, Greenwich. 2010 PIDBA (Paleoindian Database of 1995 Recent Advances and Paleoindian the Americas) 2010: Current Status and Early Archaic Period Research and Findings. Archaeology of in the Southeastern United States. Eastern North America 38:1-18. Archaeology of Eastern North America 23:145-176. Anderson, David G. Michael R. Russo, 1996 Modeling Regional Settlement in and Kenneth E. Sassaman the Archaic Period Southeast. In 2007 Mid-Holocene cultural dynamics in Archaeology of the Mid-Holocene southeastern North America. In Southeast, edited by K. E. Climate Change and Cultural Sassaman and D. G. Anderson, Dynamics: A Global Perspective on pp. 157-176. University of Florida Mid-Holocene Transitions, edited Press, Gainesville. by D. G. Anderson, K. A. Maasch 2001 Climate and Culture Change in and D. H. Sandweiss, pp. 457-490. Prehistoric and Early Historic Elsevier Press, London. Eastern North America. Archaeology of Eastern North Anderson, David G., Ashley M. America 29:143-186. Smallwood, and D. Shane Miller 2015 Pleistocene Human Settlement in Anderson, David G., and Michael K. the Southeastern United States: Faught Current Evidence and Future 2000 Paleoindian Artifact Distributions: Directions. PaleoAmerica 1:1-45. Evidence and Implications. Antiquity 74:507-513. Angst, Michael G., Bradley A. Creswell, Matthew D. Gage, Gail G. Anderson, David G. and J. Christopher Guymon, and Stephen J. Yerka Gillam 2011 Archaeological survey of TVA 2000 Paleoindian Colonization of the lands along the lower Duck River Americas: Implications from an and within the Duck River and Big Examination of Physiography, Sandy units of the Tennessee Demography, and Artifact National Wildlife Refuge, Benton, Distribution. American Antiquity Henry, and Humphreys counties, 65(1):43–66. Tennessee. Archaeological Research Laboratory, University of Anderson, David G. and Glenn T. Hanson Tennessee, Department of 1988 Early Archaic Settlement in the Anthropology. Prepared for the Southeastern United States: A Tennessee Valley Authority Case Study from the Savannah Cultural Compliance Section.

92 Colonization after Clovis

Bass, Quentin R., II Tennessee River Watersheds In In 1977 Prehistoric Settlement and the Eastern Fluted Point Tradition, Subsistence Patterns in the Great edited by J. A. M. Gingerich, pp. Smoky Mountains. M.A. thesis, 299–314. University of Utah Press, Department of Anthropology, Salt Lake City. University of Tennessee, Knoxville. Brown, James A. and Robert K. Vierra Binford, Lewis R. 1983 What Happened in the Middle 1979 Organization and Formation Archaic?: Introduction to an Processes: Looking at Curated Ecological Approach to Koster Site Technologies. Journal of Archaeology. In Archaic Hunters Anthropological Research and Gatherers in the American 35(3):255-273. Midwest, edited by J. L. Phillips 1980 Willow Smoke and Dog's Tails. and J. A. Brown, pp. 165-196. American Antiquity 45:4-20. Academic Press, New York.

Bissett, Thaddeus Buchanan, Briggs 2014 The Shell Mound Archaic in 2003 The Effects of Sample Bias on Western Tennessee: Prehistoric Paleoindian Fluted Point Recovery Occupation of the Lower in the United States. North Tennessee River Valley between American Archaeologist 24(4):311- 9000 and 2500 cal yr BP. PhD 338. dissertation, Department of Anthropology, University of Cabak, Melanie A., Kenneth E. Tennessee, Knoxville. Sassaman, and J. Christopher Gillam Brackenridge, G. Robert 1998 Distributional Archaeology in the 1984 Alluvial Stratigraphy and Aiken Plateau: Intensive Survey of Radiocarbon Dating along the E Area, Savannah River Site, Duck River, Tennessee; Aiken County, South Carolina. Implications regarding Flood-plain Savannah River Archaeological Origin. GSA Bulletin 95(1):9-25. Research Papers 8. Occasional Papers of the Savannah River Broster, John B. and Mark R. Norton Archaeological Research Program, 1996 Recent Paleoindian Research in South Carolina Institute of Tennessee. In The Paleoindian Archaeology and Anthropology, and Early Archaic Southeast, University of South Carolina, edited by D. G. Anderson and K. E. Columbia. Sassaman, pp. 288-297. University of Alabama Press, Tuscaloosa. Chapman, Jefferson and Andrea Brewer Shea Broster, John B., Mark R. Norton, D. 1981 The Archaeobotanical Record: Shane Miller, Jesse W. Tune and Early Archaic Period to Contact in Jonathan D. Baker the Lower Little Tennessee River 2013 Tennessee’s Paleoindian Record: Valley. Tennessee Anthropologist The Cumberland and Lower 6:61–84.

93 Tennessee Archaeology 8(1-2) Summer 2016

Charnov, Eric L. and Publications in Anthropology 1976 Optimal foraging: the marginal 54, Tennessee Valley Authority, value theorem. Theoretical Knoxville, Tennessee. Population Biology 9:129-136. Delcourt, Hazel R. Cramér, Harald 1979 Late Quaternary Vegetational 1946 Mathematical Methods of Statistics. History of the Eastern Highland Princeton University Press, Rim and Adjacent Cumberland Princeton. Plateau of Tennessee. Ecological Monographs 49(3):225-280. Crites, Gary D. 1985 Early Cucurbits in the Southeast: Delcourt, Paul A. and Hazel R. Delcourt New Evidence from the Hayes Site 1981 Vegetation Maps for Eastern North (40ML139). Tennessee America: 40,000 Years to Present. Anthropological Association In Geobotany: An Integrating Newsletter 10(6):8-9. Experience, edited by R. Romans. 1987 Middle and Late Holocene Plenum, New York. Ethnobotany of the Hayes Site 1983 Late Quaternary Vegetational (40ML139): Evidence from Unit Dynamics and Community Stability 990N918E. Midcontinental Journal Reconsidered. Quaternary of Archaeology 12:3-32. Research 19:263-284. 1993 Domesticated Sunflower in Fifth Millennium B.P. Temporal Context: Deter-Wolf, Aaron, Jesse W. Tune and New Evidence from Middle John B. Broster Tennessee. American Antiquity 2011 Excavations and Dating of Late 58(1):146-148. Pleistocene and Paleoindian Deposits at the Coats-Hines Site, Daniel, I. Randolph, Jr Williamson County, Tennessee. 2001 Stone Raw Material Variability and Tennessee Archaeology 5(2):142- Early Archaic Settlement in the 156. Southeastern United States. American Antiquity 66:237-265. Dunnell, Robert C 1990 The Role of the Southeast in Daniel, I. Randolph, Jr. and Albert C. American Archaeology. Goodyear Southeastern Archaeology 9:11-22. 2006 An Update on the North Carolina Fluted-Point Survey. Current Dunnell, Robert C. and William S. Dancey Research in the Pleistocene 23:99- 1983 The Siteless Survey: A Regional 101. Scale Data Collection Survey. In Advances in Archaeological Davis, Stephen Method and Theory, edited by M. 1990 Aboriginal Settlement Patterns in B. Schiffer, pp. 267-287. vol. 6. the Lower Little Tennessee River Academic Press, New York. Valley. Report of Investigation 50, Department of Anthropology, Evans, David R. University of Tennessee, Knoxville, 1972 Initial Archaeological Survey of the

94 Colonization after Clovis

Columbia Reservoir, Tennessee. Archaeology Laboratory Report submitted to the Tennessee 1983 Stop Me if You've Heard This One Valley Authority. Before: The Flint Run Paleoindian Complex Revisited. Archaeology of Faulkner, Charles H. and Major C. R. Eastern North America 11:49-59. McCollough 1973 Introductory Report of the Gill, Jacquelyn, John W. Williams, Normandy Reservoir Salvage Stephen T. Jackson, Katherine B. Project: Environmental Setting, Lininger, and Guy. S. Robinson Typology, and Survey. Department 2009 Pleistocene Megafaunal Collapse, of Anthropology, University of Novel Plant Communities, and Tennessee, Knoxville, Tennessee. Enhanced Fire Regimes in North America. Science 326:1100-1103. Fenneman, Nevin M. 1938 Physiography of the Eastern United Gillam, J. Christopher States. McGraw-Hill, New York. 1999 Paleoindian Settlement in Northeastern Arkansas. In Papers Franklin, Jay D. in Honor of Dan and Phyllis Morse, 2002 The Prehistory of Fentress County, edited by Robert C. Mainfort, Jr, Tennessee: An Archaeological ed., 99–118. University of Survey. PhD dissertation, Arkansas, Fayetteville. Department of Anthropology, University of Tennessee, Knoxville. Goodyear, Albert C. 1999 The Early Holocene Occupation of Fretwell, Steven Dewitt and Henry. L. the Southeastern United States: A Lucas, Jr. Geoarchaeological Summary. In 1970 On Territorial Behavior and Other Ice Age Peoples of North America, Factors Influencing Habitat edited by R. Bonichsen and K. L. Distribution in Birds. I. Theoretical Turnmire, pp. 432-481. Oregon Development. Acta Biotheoretica State University Press, Corvallis. 19:16-36. 2005 Evidence of Pre–Clovis Sites in the Eastern United States. In Gardner, Paul Paleoamerican Origins: Beyond 1997 The Ecological Structure and Clovis, edited by B. T. L. Robsen Behavioral Implications of Mast Bonnichsen, Dennis Stanford, and Exploitation Strategies. In People, Michael R. Waters, pp. 103-112. Plants, and Landscape: Studies in Center for the Study of the First Paleoethnobotany, edited by K. J. Americans, Texas A&M University Gremillion, pp. 161-178. University Press, College Station, Texas. of Alabama Press, Tuscaloosa. 2006 Recognizing the Redstone Fluted Point in the South Carolina Gardener, William M. Paleoindian Point Database. 1974 The Flint Run Paleoindian Current Research in the Complex: A Preliminary Report Pleistocene 23:112-114. 1971 through 1973 Seasons. Catholic University of America,

95 Tennessee Archaeology 8(1-2) Summer 2016

Graham, R. W., C.V. Haynes, D.L. Exchange Networks in the Johnson and M. Kay Southeastern United States. In 1981 Kimmswick: A Clovis Mastodon Archaeology of the Mid-Holocene Association in Eastern Missouri. Southeast, edited by K. E. Science 213:1115-1117. Sassaman and D. G. Anderson, pp. 222-234. University of Alabama Haynes, C. V. Press, Tuscaloosa. 1964 Fluted Projectile Points: Their Age 1997 Middle Archaic Bone Pins: and Dispersion. Science 145:1408- Evidence of Mid-Holocene 1413. Regional-Scale Social Groups in 2005 Clovis, Pre-Clovis, Climate the Southern Midwest. American Change, and Extinction. In Antiquity 62(3):464–487. Paleoamerican Origins: Beyond Clovis, edited by B. T. L. Robsen Jolley, Robert L. Bonnichsen, Dennis Stanford, and 1979 Archaeological Reconnaissance in Michael R. Waters, pp. 113-132. the Headwaters of the Caney Fork Center for the Study of the First River in Middle Tennessee. Americans, Texas A&M University Tennessee Anthropologist 4(1):32- Press, College Station, Texas. 62. 1980 An Archaeological Survey of the Hollenbach, Kandice D. Lower Duck and Middle 2007 Gathering in the Late Paleoindian Cumberland Rivers in Middle Period: Archaeobotanical Remains Tennessee. Manuscript on file, from Dust Cave, Alabama. In Tennessee Division of Archaeology Foragers of the Terminal Nashville, Tennessee. Pleistocene in North America, edited by B. N. Driskell and R. Kelly, Robert L and Larry C. Todd Walker, pp. 132-147. University of 1988 Coming into the Country: Early Nebraska Press, Lincoln. Paleoindian Hunting and Mobility. 2009 Foraging in the Tennessee River American Antiquity 53(2):231–244. Valley, 12,500 to 8,000 years ago. University of Alabama Press, Kelly, Robert L., Todd A. Surovell, Bryan Tuscaloosa. N. Shuman and Geoffrey M. Smith 2013 A Continuous Climatic Impact on Jefferies, Richard W. Holocene Human Populations in 1995 Late Middle Archaic Exchange and the Rocky Mountains. Proceedings Interaction in the North American of the National Academy of Midcontinent. In Native American Sciences 110:443-447. Interaction: Multiscalar Analysis and Interpretations in the Eastern Kennett, Douglas J., Atholl Anderson and Woodlands, edited by Michael S. Bruce Winterhalder Nassaney and Kenneth E. 2006 The Ideal Free Distribution, Food Sassaman, pp. 73–100. The Production, and the Colonization of University of Tennessee Press, Oceania. In Behavioral Ecology Knoxville. and the Transition to Agriculture, 1996 The Emergence of Long-Distance edited by D. J. Kennett and B.

96 Colonization after Clovis

Winterhalder, pp. 265-288. pp. 88-102. University of University of California Press, Tennessee Press, Knoxville. Berkeley. Lewis, Thomas M.N. and Madeline Kennett, Douglas J. and Bruce Kneberg Winterhalder 1959 The Archaic Culture of the Middle 2008 Demographic Expansion, South. American Antiquity Despotism, and the Colonisation of 25(2):161-183. East and South Polynesia. In Islands of Inquiry: Colonisation, Liu, Yao, Jennifer J. Anderson, John W. Seafaring, and the Archaeology of Williamson, and Stephen T. Maritime Landscapes, edited by G. Jackson Clark, F. Leach and S. O'Connor, 2013 Vegetation History in Central pp. 87-96. Australia National Kentucky and Tennessee (USA) University Press, Canberra. during the Last Glacial and Deglacial Periods. Quaternary Kerr, Jonathan P. and Andrew Bradbury Research 79:189-198. 1998 Paleo-indian and Early Archaic Settlement at Kentucky Lake. Maggard, Greg and Kacy Stackelbeck Tennessee Anthropologist 23(1- 2008 Paleoindian Period. In The 2):1-20. Archaeology of Kentucky: An Update - Volume One, edited by Kimball, Larry R. David Pollack, pp. 109-192. 1996 Early Archaic Settlement and Preservation Plan Report No. 3. Technology: Lessons from Tellico. Kentucky Heritage Council State In The Paleoindian and Early Historic, Frankfurt Archaic Southeast, edited by D. G. Anderson and K. E. Sassaman, pp. Martin, Paul S 149-186. University of Alabama 1973 The discovery of America. Science Press, Tuscaloosa. 179:969–74. 2006 Twilight of the Mammoths: Ice Age Klippel, Walter E. and Paul W. Parmalee Extinctions and the Rewilding of 1982 Diachronic Variation in Insectivores America. Berkeley: University of from Cheek Bend Cave and California Press. Environmental Change in the Midsouth. Paleobiology 8(4):447- Mason, Ronald J. 458. 1962 The Paleo-Indian Tradition in Eastern North America. Current Lane, Leon, and David G. Anderson Anthropology 3:227-228. 2001 Paleoindian Occupations of the Southern Appalachians: A View McClure, Sarah B., Michael A. Jochim, from the Cumberland Plateau of and C. Michael Barton Kentucky and Tennessee. In 2006 Human Behavioral Ecology, Archaeology of the Appalachian Domestic Animals, and Land Use Highlands, edited by Lynne P. during the Transition to Agriculture Sullivan and Susan C. Prezzano, in Valencia, Eastern Spain. In

97 Tennessee Archaeology 8(1-2) Summer 2016

Behavioral Ecology and the Ecological Analysis of Paleoindian Transition to Agriculture, edited by and Archaic Subsistence and D. J. Kennett and B. Winterhalder. Landscape Use in Central University of California Press, Tennessee. PhD dissertation, Berkeley. Department of Anthropology, University of Arizona, Tucson. McNutt, Charles H. 2008 The Benton Phenomenon and Miller, D. Shane and Joseph A.M. Middle Archaic Chronology in Gingerich Adjacent Portions of Tennessee, 2013 Regional Variation in the Terminal Mississippi, and Alabama. Pleistocene and Early Holocene Southeastern Archaeology Radiocarbon Record of Eastern 27(1):45-60. North America. Quaternary Research 79:175-188. McNutt, Charles H. John B. Broster and Mark R. Norton Mills, H.H. and Paul A. Delcourt 2008 A Surface Collection from the Kirk 1991 Quaternary Geology of the Type Site (40HS174), Humphreys Appalachian Highlands and Interior County, Tennessee. Tennessee Low Plateaus. In Quaternary Non- Archaeology 3(1):25-75. Glacial Geology: Conterminous U.S., edited by R. B. Morrison, pp. Meltzer, David J. 611-628. Geology of North 1988 Late Pleistocene Human America. Vol. K-2. Geological Adaptations in Eastern North Society of America, Boulder. America. Journal of World Prehistory 2:1-53. Milner, George R. 2004 Modeling the Initial Colonization of 1999 Warfare in Prehistoric and Early the Americas: Issues of Scale, Historic Eastern North America. Demography, and Landscape Journal of Archaeological Learning. In The Settlement of the Research 7(2):105-151. American Continents: A Multidisciplinary Approach to Morse, Dan F. Human Biogeography, edited by C. 1971 Recent Indications of Dalton Michael Barton, Geoffrey A. Clark, Settlement Pattern in Northeast David R. Yesner, and Georges A. Arkansas. Southeastern Pearson, pp. 123-137. The Archaeological Conference Bulletin University of Arizona Press, 13: 5–10. Tucson. 1973 Dalton Culture in Northeast 2009 The First Peoples in a New World: Arkansas. Florida Anthropologist Colonizing Ice Age America. 26(1):23-28. University of California Press, 1975 Reply to Schiffer. In The Cache Berkeley. River Archaeological Project: An Experiment in Contract Miller, D. Shane Archaeology, edited by M. B. 2014 From Colonization to Schiffer, and J. M. House, pp. 113– Domestication: A Historical 19. Research Series 8. Arkansas

98 Colonization after Clovis

Archaeological Survey, Fayetteville. Fayetteville. 1975b An Alternative to Morse’s Dalton 1977 Dalton settlement systems: Reply Settlement Pattern Hypothesis. to Schiffer (2). Plains Plains Anthropologist 20: 253–66. Anthropologist 22: 149–58. Shennan, Stephen and Kevan Munson, Patrick J. Edinborough 1986 Hickory Silviculture: A Subsistence 2007 Prehistoric Population History from Revolution in the Prehistory of the Late Glacial to the Late Eastern North America. Paper Neolithic in Central and Northern presented at Emergent Europe. Journal of Archaeological Horticultural Economies of the Science 34:1339-1345. Eastern Woodlands Conference, Southern Illinois University, Shott, Michael J. Carbondale. 2002 Sample Bias in the Distribution and Abundance of Midwestern Fluted Nelson, Margaret Bifaces. Midcontinental Journal of 1991 The Study of Technological Archaeology 27:89-123. Organization. Archaeological 2005 Representivity of the Midwestern Method and Theory 3:57-100. Paleoindian Site Sample. North American Archaeologist 25:189- Pearson, Karl 212. 1900 On the Criterion that a Given System of Deviations from the Smallwood, Ashley M. Probable in the Case of a 2012 Clovis Technology and Settlement Correlated System of Variables Is in the American Southeast Using Such That It Can Be Reasonably Biface Analysis to Evaluate Supposed to Have Arisen from Dispersal Models. American Random Sampling. Philosophical Antiquity 77(4):689-713. Magazine 50(302):157-175. Smith, Bruce D. Prasciunas, Mary M. 1987 Independent Domestication of 2011 Mapping Clovis: Projectile Points, Seed Bearing Plants in Eastern Behavior, and Bias. American North America. In Emergent Antiquity 76(1):107-126. Horticultural Economies of the Eastern Woodlands, edited by W. Schiffer, Michael B F. Keegan, pp. 3-47. Occasional 1975a Some Further Comments on the Paper No. 7, Center for Dalton Settlement Pattern Archaeological Investigations, Hypothesis. In The Cache River Southern Illinois University, Archaeological Project: An Carbondale. Experiment in Contract 1992 The Floodplain Weed Theory of Archaeology, edited by M. B. Plant Domestication in Eastern Schiffer, and J. H. House, pp. 103– North America. In Rivers of 12. Research Series No. 8. Change: Essays on Early Arkansas Archaeological Survey, Agriculture in Eastern North

99 Tennessee Archaeology 8(1-2) Summer 2016

America, edited by B. D. Smith, pp. Accomplishments and Future 19-34. Smithsonian Institution Directions, edited by D. Pollack, Press, Washington, D.C. pp. 73–142. Kentucky Heritage 2001 Low Level Food Production. Council, Frankfurt. Journal of Archaeological Research 9(1):1-43. Tankersley, Kenneth B., Michael R. 2007 Niche Construction and the Waters, and Thomas W. Stafford, Behavioral Context of Plant and Jr. Animal Domestication. Evolutionary 2009 Clovis and the American Mastodon Anthropology 16(5):188-199. at Big Bone Lick, Kentucky. 2011 The Cultural Context of Plant American Antiquity 74(3): 558-567. Domestication in Eastern North America. Current Anthropology Thomas, David Hurst 52(S4):S471-S484. 1975 Nonsite Sampling in Archaeology: Up the Creek Without a Site. In Smith, Maria O. Sampling in Archaeology, edited by 1995 Scalping in the Archaic Period: J. W. Mueller, pp. 61-81. University Evidence from the Western Valley. of Arizona Press, Tucson, Arizona. Southeastern Archaeology 14(1):60-68. Thulman, David 1996 Bioarchaeological Inquiry into 2009 Freshwater Availability as the Archaic Period Populations of the Constraining Factor in the Middle Southeast: Trauma and Paleoindian Occupation of North- Occupational Stress. In The Mid- Central Florida. Geoarchaeology Holocene Southeast, edited by K. 24(3):243-276. E. Sassaman and D. G. Anderson, pp. 134-154. University of Alabama United States Geologic Survey (USGS) Press, Tuscaloosa. 2013 National Elevation Dataset. Electronic database, Surovell, Todd A. http://ned.usgs.gov, accessed 2009 Toward a Behavioral Ecology of March 2014. Lithic Technology: Cases from Paleoindian Archaeology. Viau, A.E., K. Gajewski, M.C. Sawada University of Arizona Press, and P. Fines Tucson. 2006 Millennial-scale Temperature Variations in North America during Surovell, T. A., J. Finely, G. M. Smith, P. the Holocene. Journal of J. Brantingham and R. L. Kelly Geophysical Research 111(D9):1- 2009 Correcting Temporal Frequency 12. Distributions for Taphonomic Bias. Journal of Archaeological Science Walker, Renee 36:1715-1724. 1998 The Late Paleoindian through Middle Archaic Evidence from Dust Tankersley, Kenneth B. Cave, Alabama. PhD dissertation, 1990 Paleoindian Period. In The Department of Anthropology, Archaeology of Kentucky: Past University of Tennessee, Knoxville.

100 Colonization after Clovis

2007 Hunting in the Late Paleoindian Period: Faunal Remains from Dust Cave, Alabama. In Foragers of the Terminal Pleistocene, edited by R. B. Walker and B. N. Driskell, pp. 99-115. University of Alabama Press, Tuscaloosa.

Walthall, John A. 1998 Rockshelters and Hunter-Gatherer Adaptation to the Pleistocene/ Holocene Transition. American Antiquity 63(2):223-238

Wilkins, Gary R., Paul A. Delcourt, Hazel R. Delcourt, Frederick W. Harrison and Manson R. Turner 1991 Paleoecology of Central Kentucky since the Last Glacial Maximum. Quaternary Research 36(2):224- 239.

Williams, John W., Bryan N. Shuman, Thompson III Webb, Patrick J. Bartlein and Phillip Ledug 2004 Late-Quaternary Vegetation Dynamics in North America: Scaling from Taxa to Biomes. Ecological Monographs 74(2):309- 334.

Winterhalder, Bruce, Douglas J. Kennett, Mark N. Grote and Jacob Bartrund 2010 Ideal Free Settlement Distribution of California's Northern Channel Islands. Journal of Anthropological Archaeology 29:469-490.

D. Shane Miller Dept of Anthropology and Middle Eastern Cultures Mississippi State University [email protected]

Stephen B. Carmody Department of Earth and Environmental Systems Sewanee: University of the South [email protected]

101 THE PALEOINDIAN AND EARLY ARCHAIC HILLTOP OCCUPATIONS AT THE TOPPER SITE

Derek T. Anderson, Ashley M. Smallwood, Albert C. Goodyear, and Sarah E. Walters

Recent AMS dating of charred remains from the Paleoindian occupation of the upper hillside area at the Topper site has provided the first precise radiocarbon date in the Southeast that is directly associated with diagnostic Clovis lithic artifacts. This paper presents the results of dating, geoarchaeological, and lithic analyses in this area of the site, with a focus on the Paleoindian and Early Archaic components.

Topper (38AL23) is well-known as a Woodland, and Mississippian lithic quarry-related site, with an extensive assemblages (Goodyear 2001). In 2009, Clovis component (Goodyear and Charles the first of a series of units were 1984; Goodyear and Steffy 2003), but excavated on the upland portion of the little attention has been paid to the post- site (referred to as the “hilltop”), in order to Clovis utilization of the site (Goodyear determine the spatial extent of the Clovis 2001; Miller 2010:25-33; Sain 2012:47). material, and excavations have continued Likewise, while the production, use, and on the hilltop on a limited basis since discard of stone tools at Topper has been then. This article describes some of the the focus of multiple analyses (Goodyear recent (2009-2014) investigations of the and Steffy 2003; Goodyear et al. 2009; Topper site, with a focus on the Miller and Goodyear 2008; Miller and Paleoindian and Early Archaic Smallwood 2012; Sain 2012; Smallwood assemblages, and the establishment of a 2010; 2011; Smallwood and Goodyear cultural chronology on the hilltop. 2009; Steffy and Goodyear 2006), unmodified debitage has received very Hilltop Excavations little attention. From 1998 to 2008, annual fieldwork Excavations on the hilltop began as a at Topper focused on the terrace and series of 2x2 m “test units” placed hillside portions of the site (Figure 1). progressively farther from the river and Excavations of large, contiguous blocks the chert outcrop, and the most productive as well as scattered 2x2 m units across of these initial units were expanded into the site resulted in the recovery of four three contiguous blocks. Terrestrially- finished and discarded Clovis point bases outcropping nodules are available about (Smallwood 2011:75), numerous early, 100 meters from the excavation block; middle, and late stage Clovis preforms nodules with “river cortex” or “river with evidence of end thinning and staining” can be found in and along the overshot flaking, and many categories of banks of the Savannah River, about 200 other formal tools (Smallwood et al. meters away. Accumulations of heavily 2013). Early Archaic Taylor points, tools, weathered debitage (e.g., Miller 2010:35) and debitage overlie the Clovis material in lacking diagnostic bifacial tools have been many areas of the site, and these are encountered in all units on the hilltop, in deposited below Middle and Late Archaic, some cases at depths greater than one

102 Topper Site

FIGURE 1. Map of excavation units at the Topper site with terrace, hillside, and hilltop areas highlighted. Adapted from Miller (2010).

meter below ground surface. Although outcrop. Three units forming an “L” shape excavations on the hilltop began in 2009, were excavated in 2009, and a fourth unit demonstrating this was a thoroughly that extended the block an additional two utilized area of the site beginning at least meters to the north was added in 2012. in the Early Archaic period, no diagnostic Two Early Archaic side-notched points Clovis artifacts were recovered until the (Figure 2) were found associated with an 2011 season. Because multiple accumulation of debitage at about 80 cm components are likely represented in below ground surface; a variety of non- each of the three blocks, each block has diagnostic bifacial and flake tools been assigned a sequential numeric including a large macroblade and an end- designation in order to differentiate them thinned or fluted preform were found up to from large excavation blocks in other 15 cm below the Early Archaic material parts of the site. and likely represent a separate, earlier component. No detailed analysis of this Block I material has occurred at this point, but sediment samples have been collected for The first contiguous block excavated flotation and particle size analysis. on the hilltop is also closest to the chert

103 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 2. Early Archaic side-notched points FIGURE 3. Kirk Corner-notched point recovered in Block I. recovered in Block II.

Block II debitage and charred botanical material was recovered to a depth of In 2010, a 2x2 m unit was opened approximately 130 cm below surface in about 30 meters north of Block I. At both of these units as well, but no approximately 109 cm below ground diagnostic tools were recovered below the surface, a Kirk corner-notched point was Kirk component. recovered (Figure 3) associated with a lithic scatter and a faint lens of ash and Block III charred botanical remains that likely represent a hearth or cooking feature. The A second 2x2 m unit was initiated in unit was re-opened in 2011, and 2010, about 25 meters east of Block II. excavations continued to a depth of 130 Excavations were halted at the end of the cm below surface. Although a small season before the unit could be amount of cultural material was present at completed, due to an extremely dense this depth, and the unit never fully layer of weathered lithic material “bottomed out,” excavations were halted consisting of debitage and non-diagnostic for safety reasons and no diagnostic tools tools a meter below ground surface. In were found below the Kirk component. In 2011, the unit was re-opened, and the 2012, two additional 2x2 m units were discovery of Clovis artifacts including a excavated adjacent to the original unit. late stage fluted preform, an overshot The unit to the north contained a large flake, and multiple tools (including disturbed area and was only partially macroblade fragments) to a depth of 115 excavated, but the unit to the west was cm below ground surface led to the fully excavated and contained excavation of an adjacent 2x2 m unit to endscrapers, bifacial tools or preforms, the east. Over 600 artifacts were mapped and flake tools at the same 5 cm level as in situ in a 15 cm “lithic floor” in the the Kirk point and hearth. Scattered second unit, so two additional 2x2 m units

104 Topper Site

Table 1. Counts of Bifaces with Evidence of Thinning Flake Scars from the Paleoindian Levels of Block III.

FIGURE 4. Bifaces from the Paleoindian component in Block III. were excavated to the north in 2012, remove cortex and irregularities. After resulting in a 4x4m block. initial reduction, bifaces were thinned and shaped. Four bifaces retain overshot flake Hilltop Assemblage scars, 20 bifaces have evidence of broad overface removals, and 13 bifaces retain Twenty-seven bifaces and biface scars from end thinning (Table 1). fragments were recovered from the buried Preforms generally have squared, Paleoindian component (Figure 4). Of beveled bases and excurvate lateral these, four fragments mend as refitted margins. Debitage that can be considered bifaces, making three complete bifaces diagnostic of Clovis reduction (Bradley et and 22 fragments. Based on al. 2010; Smallwood 2012; Waters et al. morphological characteristics, shape 2011; also see Eren et al. 2013) was also ratios, and the flaking index of reduction recovered, including one complete (Miller and Smallwood 2012; Smallwood overshot flake, four distal portions of 2010), the hilltop assemblage has 11 overshot flakes, and two end thinning early stage bifaces, six middle stage flakes. A single Dalton (see Goodyear bifaces, and eight late stage bifaces. 1974) preform made of local Coastal Plain Hilltop bifaces were crafted from large chert (Figure 5) was recovered in the spalls struck from locally available chert upper portion of the Paleoindian nodules. Spalls were bifacially flaked to component as well. Its collateral flaking to

105 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 5. Dalton preform recovered from the Paleoindian component in Block III. the midline, lobed ears, and planview greater than 0.70, which are values distinguish it from Clovis, and it appeared experimentally shown by Kuhn to exhibit to have been pushed down into the retouch resulting from greater than 5 deposits, as it was resting at a 60 degree reduction events. slope. This late Paleoindian point is Based on the presence of bifacial noteworthy as it is only the second post- reduction debitage and the ratio of bifaces Clovis, diagnostic Paleoindian artifact per square meter (1.75), the hilltop served found in over 800 square meters of as a biface production locus, like other excavations at Topper; the other is an areas of the Topper site. However, the orthoquartzite Redstone point from the hilltop is unique in the greater density of hillside (Miller 2010:31). other tool types. Based on simple The hilltop Paleoindian assemblage comparisons of density per square meter, also includes 21 formal flake and core the density of flake/core tools and tools and at least 15 retouched flakes. Of modified flakes in the Paleoindian the flake and core tools, there are eight component is 2.25 tools per square meter sidescrapers, seven endscrapers, two for Block III, compared to the 0.68 value cobble chopping tools, one denticulate, calculated for the 128 square meters of one adze preform, one planer blank, and the hillside assemblage (Smallwood et al. one wedge tool (following Andrefky 2005; 2013). Clearly, flake and core tool also see Smallwood et al. 2013 for tool production and use were important type descriptions for Topper). Based on activities taking place in the hilltop portion Kuhn’s (1990) calculation of Reduction of the site. While these activities are Index (RI), six of these tools have RIs evident in other areas of Topper,

106 Topper Site

Lithic Refitting

In 2010, a 4x4m block was excavated on the terrace that contained a mixed Paleoindian and Taylor assemblage. A detailed analysis of debitage from the block ultimately resulted in the refitting of over 25% of the mapped artifacts, providing a means of differentiating between the two components and demonstrating that the deposits at Topper are relatively undisturbed, both horizontally and vertically (Anderson 2011). Although no charcoal, staining, or textural differences were observed during FIGURE 6. Plan view of a hearth-centered excavations, analysis of individually activity area in the "terrace block" at Topper. mapped items shows that thermally Black lines represent refits; red diamonds are altered and burned artifacts are thermally altered artifacts, and black squares concentrated in two areas within the are unaltered artifacts. Hearth-centered block. When the refit distributions are seating model via Binford 1978. overlaid, these two areas are recognizable as activity areas, likely hearths (Figure 6). Refitting in the hilltop excavation blocks has demonstrated that other portions of the site are equally well- preserved and intact. Preliminary analysis of the lithics from a portion of Block II (Figure 7) indicates that at least one hearth-centered activity area exists in the northern portion of the block, in addition to the hearth that was bisected in the southern wall. Average vertical distance between all refits in Block II is 3.8 cm, and about 7.5% of the mapped assemblage from the two units that have been analyzed so far refits (86 of 1147 artifacts). Refitting of the third 2x2 m unit FIGURE 7. Preliminary results of lithic refitting in this block is planned and should help to from Block II. Shaded portion represents increase the overall percentage of refits disturbed area that was not fully excavated. within the assemblage. Detailed spatial analysis and refitting especially the hillside (Smallwood 2015), of Block III is also in progress and has the hilltop has produced the densest produced similar results. Concentrations evidence of flake and core tool production of artifacts that likely represent knapping and use thus far. clusters and hearths are visible in

107 Tennessee Archaeology 8(1-2) Summer 2016

(Hollenbach 2009:98). The Flamingo Bay site (38AK469) has produced a series of dates on Early Archaic material (Moore et al. 2012), but no other published dates exist for the late Pleistocene or early Holocene in South Carolina (Miller and Gingerich 2013a, 2013b), severely limiting discussions of regional chronologies and culture change. Historically, attempts to date the Clovis occupation at Topper have been relatively unsuccessful. Waters et al. (2009:1305) were only able to collect one radiocarbon sample from the Clovis deposits on the terrace, but it resulted in a date of 2,170 +/- 40 B.P. (CAMS-66110; charcoal) and was clearly introduced from above. FIGURE 8. Preliminary results of lithic refitting Although a series of optically stimulated from Block III. luminescence dates more accurately planview (Figure 8), and although more dated Clovis at the site, they were also vertical movement of material is present in considered unreliable due to bioturbation, this part of the site (tree roots or an and standard deviations of over 1,000 intrusive pit appear to have moved many years make them unsuitable for isolating artifacts in the northeast corner of the individual cultural components (Waters et block), many features seem to be largely al. 2009:1308). intact. The success rate for refitting in Beginning in 2010, however, Block III is currently over 13% of the excavations on the hilltop have resulted in mapped assemblage (253 of 1,918 the recovery of large quantities of small, artifacts), and the excavation and analysis charred botanical remains both in situ and of additional units in the future should aid from the 1/8” screens. After recognizing in our understanding of late Pleistocene that charred organic remains in Block II and early Holocene spatial organization were likely associated with an ancient among hunter-gatherer groups. hearth feature, a series of test columns were excavated across the hilltop for Dating and Chronology at Topper flotation analysis. Presence/absence studies and preliminary analysis from one The acidic soils that are common to of the columns identified a variety of many archaeological sites in the paleobotanical remains, many of which southeastern United States are often are likely culturally deposited (Walters devoid of recoverable organic remains 2013; Walters et al. 2013). This suggests (Hollenbach 2009:1), and the sandy that botanical remains are not only deposits of the Coastal Plain are no present throughout the hilltop portion of exception. As a result, radiocarbon dates the site, but are found in all cultural from the Paleoindian and Early Archaic deposits, from the Paleoindian period to periods are relatively rare in the the present. Since 2011, a concerted Southeast, particularly at open-air sites effort has been made to collect as many

108 Topper Site

Table 2. Radiocarbon Samples from the Topper Hilltop Units. All Dates are Reported in Uncalibrated Calendar Years.

samples of organic material as possible Walters et al. 2013) and falls within from the excavation units. Dozens of Waters and Stafford’s (2007) redefined pieces of charred material have been range for the Clovis period (11,500 - mapped in situ, and hundreds more have 10,900 BP; 13,250-12,800 CALYBP). Two been recovered from the screens. additional softwood samples were then In 2011, a sample of hickory nut shell selected for dating from material associated with the Kirk corner-notched recovered in the 1/8” screens from point in Block II was dated at 8,130 +/- 40 adjacent units. A roughly 2,100 year-old BP (Beta-296974, delta 13C = -23.3) sample likely represents a relatively indicating that old charcoal existed on the recent root burn, but the second falls hillside and providing another line of within the Younger Dryas chronozone evidence that minimal post-depositional (12,850-11,700 cal yr BP, a time span that movement of cultural material had is under-represented in the Southeast occurred within the deposits. After [Miller and Gingerich 2013a]) and could successfully dating the Kirk component, date a late Dalton or early Taylor three additional samples from the same occupation at the site. excavation block were selected and submitted to radiocarbon dating labs in Conclusion 2012 and 2013. A sample of black gum was clearly displaced from above (3,306 Topper represents a unique and +/- 41 BP, AA-100292, delta 13C = -23.5), persistent place on the landscape but a piece of muscadine at the base of (Schlanger 1992) that was utilized by the Kirk occupation corroborates the initial most, if not all, of the various groups of Kirk date, and a diffuse-porous hardwood people living on the Coastal Plain of associated with the basal lithic floor at South Carolina for the past 13,000 years. approximately 130 cm below surface likely Like most of the sandy sites in the region, dates the Taylor component on the hilltop Topper does not have clearly defined (Table 2). stratigraphic breaks that make it possible Three samples from Block III were to pick a spot on the profile wall of a unit also dated in 2013 (Table 2). A cold- and determine its age based on depth adapted softwood (spruce, fir, or larch) alone. However, Topper does have a from one of Walters’ flotation columns large number of diagnostic artifacts, provided the first - and thus far, only - surprisingly well-preserved botanical radiocarbon date from the Clovis remains, and a steadily-increasing component at the site (Goodyear 2013; number of dates. Based on the current

109 Tennessee Archaeology 8(1-2) Summer 2016

results, it appears the Kirk component Acknowledgements. Thanks to Jesse Tune and dates from roughly 8000-9000 BP, Taylor Shane Miller for organizing the SEAC session in 2013 and asking us to participate. Kandace from 9000-10,000 BP, and the Dalton Hollenbach provided identifications for many of the through Clovis components from 10,000- charred remains that have been recovered. Thank 11,000 BP, paralleling discoveries you to all of the Topper volunteers, staff, and elsewhere in the Southeast. Additional supervisors who have spent so much time radiocarbon and OSL samples are excavating and mapping the units, bagging artifacts, and filling out paperwork – we would currently being processed, and with the have nothing without your hard work and continued recovery of carbonized dedication. We also thank Leslie Page, Joan and remains, Topper has the potential to not Ernie Plummer, and an anonymous donor for only securely date Clovis in the paying for the initial radiocarbon dates, as well as Southeast, but to construct a complete all of the donors who support SCIAA, SEPAS, and the Cobb Institute. Finally, thank you to Clariant cultural chronology for the region as well. and Archroma for continued access to and Different occupational episodes at the site preservation of the site and for the use of the can also be identified through the use of a campground and facilities on the property. variety of analytical techniques. Refitting in all three hilltop blocks, along with References minimal nodule analysis (MANA) (e.g., Larson and Kornfeld 1997) and Anderson, Derek T. microdebitage analysis continue to help 2011 The 2010 4x4 Meter Unit at differentiate and define the assemblages Topper: Preliminary Lithic and left at the site by the prehistoric people Spatial Analyses. Legacy who visited Topper. (Newsletter of the South Carolina Further, it is becoming increasingly Institute of Archaeology and clear that the Paleoindian occupation at Anthropology) 15(1):16-19. Topper is complex and spatially varied. With the presence of overshot flakes, a Andrefsky, William Jr. debitage type thus far shown to be 2005 Lithics: Macroscopic Approaches to diagnostic of Clovis, the Dalton point Analysis. Second Edition. preform, and an Early Archaic Cambridge University Press, radiocarbon date, the expansion of Block Cambridge. III on the hilltop in particular may be key to revealing the nature of a post-Clovis Bradley, B. A., M. B. Collins, and A. Paleoindian occupation at Topper, which Hemmings outside of this block thus far appears to 2010 Clovis Technology. Archaeological be sparse. If further analysis of the hilltop Series 17, International proves that Topper was extensively used Monographs in Prehistory. by Dalton populations as a workshop and University of Michigan, Ann Arbor. camp, like the Clovis occupation has shown, this would be the first recorded Eren, M. I., R. J. Patten, Michael J. Late Paleoindian-age quarry-related site O’Brien, and D. J. Meltzer in the Southeast. Certainly, this would 2013 Refuting the Technological provide a significant opportunity to fully Cornerstone of the Ice-Age Atlantic understand post-Clovis production Crossing Hypothesis. Journal of technology and settlement. Archaeological Science 40: 2934- 2941.

110 Topper Site

Goodyear, Albert C. Archaeological Science 17:585- 1974 The Brand Site: A Techno- 593. functional Study of a Dalton Site in Northeast Arkansas. Research Larson, Mary Lou and Marcel Kornfeld Series 7. Fayetteville: Arkansas 1997 Chipped Stone Nodules: Theory, Archaeological Survey. Method, and Examples. Lithic 2001 The 2001 Allendale Paleoindian Technology 22:4-18. Expedition and Beyond. Legacy (Newsletter of the South Carolina Miller, D. Shane Institute of Archaeology and 2010 Clovis Excavations at Topper Anthropology) 6(2):18-21. 2005-2007: Examining Site 2013 Update on the 2012-2013 Activities Formation Processes at an Upland of the Southeastern Paleoamerican Paleoindian Site along the Middle Survey. Legacy (Newsletter of the Savannah River. Occasional Paper South Carolina Institute of No. 1 of the Southeastern Archaeology and Anthropology) Paleoamerican Survey, South 17(1):10-12. Carolina Institute of Archaeology and Anthropology, Columbia. Goodyear, Albert C. and Tommy Charles 1984 An Archaeological Survey of Chert Miller, D. Shane and Joseph A. M. Quarries in Western Allendale Gingerich County, South Carolina. University 2013a Regional Variation in the Terminal of South Carolina, Institute of Pleistocene and Early Holocene Archaeology and Anthropology, Radiocarbon Record of Eastern Research Manuscript Series 195, North America. Quaternary Columbia, SC. Research 79:175-188. 2013b Paleoindian Chronology and the Goodyear, Albert C., Keith Derting, D. Eastern Fluted Point Tradition. In Shane Miller, and Ashley M. The Eastern Fluted Point Tradition, Smallwood pp. 9-37, edited by Joseph A. M. 2009 Exotic Clovis Stone Tools from the Gingerich. University of Utah Topper Site, 38AL23, Allendale Press, Salt Lake City. County, South Carolina. Current Research in the Pleistocene 26:60- Miller, D. Shane and Albert C. Goodyear 62. 2008 A Probable Hafted Uniface from the Clovis Occupation at the Goodyear, Albert C. and Kenn Steffy Topper Site, 38AL23, Allendale 2003 Evidence of a Clovis Occupation at County, South Carolina. Current the Topper Site, 38AL23, Allendale Research in the Pleistocene County, South Carolina. Current 25:118-120. Research in the Pleistocene 20:23- 25. Miller, D. Shane and Ashley M. Smallwood Kuhn, Steven 2012 Beyond Stages: Modeling Clovis 1990 A Geometric Index of Reduction for Biface Production at the Topper Unifacial Stone Tools. Journal of Site (38AL23), South Carolina. In

111 Tennessee Archaeology 8(1-2) Summer 2016

Contemporary Lithic Analysis in the 2425. Southeast: Problems, Solutions, 2011 Clovis Technology and Settlement and Interpretations, pp. 28-41, in the American Southeast. edited by Phil Carr, Andrew Unpublished PhD dissertation, Bradbury, and Sarah Price. Department of Anthropology, University of Alabama Press, Texas A&M University. Tuscaloosa. 2012 Clovis Technology and Settlement in the American Southeast Using Moore, Christopher R., Mark J. Brooks, Biface Analysis to Evaluate Andrew H. Ivester, Terry Ferguson, Dispersal Models. American and James K. Feathers Antiquity 77: 689–713. 2012 Radiocarbon and Luminescence 2015 An Evaluation of Assemblage Dating at Flamingo Bay (38AK469): Context Formation and Spatial Implications for Site Formation Organization at Topper, a Clovis Processes and Artifact Burial at a Site in the American Southeast. Carolina Bay. Legacy (Newsletter Journal of Field Archaeology. In of the South Carolina Institute of press, January 2015. Archaeology and Anthropology) 16(1):16-21. Smallwood, Ashley M. and Albert C. Goodyear Sain, Douglas A. 2009 Reworked Clovis Biface Distal 2012 Clovis Blade Technology at the Fragments from the Topper Site, Topper Site (38AL23): Assessing 38AL23: Implications for Clovis Lithic Attribute Variation and Technological Organization in the Regional Patterns of Technological Central Savannah River Region. Organization. Occasional Paper Current Research in the No. 2 of the Southeastern Pleistocene 26:118-120. Paleoamerican Survey, South Carolina Institute of Archaeology Smallwood, Ashley M., D. Shane Miller, and Anthropology, Columbia. and Douglas Sain 2013 An Overview of the Clovis Lithic Schlanger, Sarah H. Assemblage from the Topper 1992 Recognizing Persistent Places in Hillside. In In The Eastern Fluted Anasazi Settlement Systems. In Point Tradition, pp. 280-298, edited Space, Time, and Archaeological by Joseph A. M. Gingerich. Landscapes, pp. 91-112, edited by University of Utah Press, Salt Lake Jaqueline Rossignol and LuAnn City. Wandsnider. Plenum Press, New York. Steffy, Kenn and Albert C. Goodyear 2006 Clovis Macro Blades from the Smallwood, Ashley M. Topper Site, 38AL23, Allendale 2010 Clovis Biface Technology at the County, South Carolina. Current Topper site, South Carolina: Research in the Pleistocene Evidence for Variation and 23:147-149. Technological Flexibility. Journal of Archaeological Science 37:2413-

112 Topper Site

Walters, Sarah E. Derek T. Anderson 2013 Tell Us What You Ate – and We’ll Cobb Institute of Archaeology PO Box AR Tell You Who, What, and When Mississippi State, MS 39762 You Were: Paleoethnobotanical [email protected] Recovery and Analysis along the Savannah River. Paper presented Ashley M. Smallwood at the 79th annual meeting of the Department of Anthropology Antonio J. Waring, Jr. Archaeology Laboratory Society for American Archaeology, University of West Georgia Austin, Texas. 1601 Maple Street Carrollton, GA 30118 Walters, Sarah E., Sean Cary von Gunter, [email protected] and Albert C. Goodyear Albert C. Goodyear 2013 The Secret Life of Carbonized South Carolina Institute of Archaeology and Plant Remains: A Guide to Dating Anthropology in the Age of Clovis. Poster University of South Carolina presented at the Paleoamerican 1321 Pendleton St, Columbia, SC 29208 Odyssey conference, Santa Fe, [email protected]

New Mexico. Sarah E. Walters Department of Anthropology Waters, Michael R., Steven L. Forman, University of Tennessee Thomas W. Stafford, Jr., and John 250 South Stadium Hall Foss Knoxville, Tennessee, 37996-0720 [email protected] 2009 Geoarchaeological Investigations at the Topper and Big Pine Tree Sites, Allendale County, South Carolina. Journal of Archaeological Science 36:1300-1311.

Waters, M. R., C. D. Pevny, D. L. Carlson, W. A. Dickens, A. M. Smallwood, S. A. Minchak, E. Bartelink, J. M. Wiersema, J. E. Wiederhold, H. M. Luchsinger, D. A. Alexander, and T. A. Jennings 2011 A Clovis Workshop in Central Texas: Archaeological Investigations of Excavation Area 8 at the Gault Site. Texas A&M University Press, College Station.

Waters, Michael R. and Thomas W. Stafford, Jr. 2007 Redefining the Age of Clovis: Implications for the Peopling of the Americas. Science 315 (5815):1122-1126.

113 CLOVIS BLADE TECHNOLOGY AND TOOL USE ALONG THE SOUTH ATLANTIC COASTAL PLAIN AND PIEDMONT OF THE LOWER SOUTHEAST

Douglas Sain and Albert C. Goodyear

John Broster has contributed to our knowledge and understanding of Southeastern Archaeology over the course of his career, and in this capacity has been an inspiration to the work of numerous individuals. The discovery of Paleoindian sites across the Southeastern U.S. has revealed a substantial presence of blades, and blade cores. While fluted points have been extensively recorded as part of the Paleoindian Database of the Americas, less research has considered the role of blades in such contexts. This paper presents the formation and development of a Clovis blade database to account for the distribution of these artifacts across the Southeast U.S. A technological and morphological analysis of a sample of blades from this dataset demonstrates patterns of variation when compared with blades from known quarry sites in the Central Savannah River Valley.

The discovery of Paleoindian sites and Big Pine Tree sites, two known quarry across the Southeast U.S. has revealed a related lithic manufacture sites in the substantial presence of chipped stone Central Savannah River Valley that form a tools including fluted points, prismatic part of what has been referred to as the blades, and blade cores. Fluted points Allendale Brier Creek Complex (Sain and have been extensively recorded as part of Goodyear 2012). The morphological and the Paleoindian Database of the technological attributes of these blades Americas, with data available on nearly were examined to explore patterns of 30,000 artifacts to date (Anderson et al. artifact variation across space. The results 2010). However, less research has of this study demonstrate that modified considered the distribution and role of blades recovered from isolated localities Clovis prismatic blades across similar of the Coastal Plain of and Piedmont contexts. Blades, as a form of stone tool South Carolina and Georgia are technology, are increasingly being technologically similar to blades recognized as an integral element of the recovered from identified quarry sites Clovis lithic toolkit (Bradley et al. 2010). within the Central Savannah River Valley. The comparison of blades from excavated However, morphological differences exist and isolated contexts throughout the which suggest that blades were being Southeast U.S. provides an excellent treated differently once transported away opportunity to examine the role of blades from these quarries. One byproduct of this in the organization of Clovis life-ways. study is the development of a regional This paper presents the results of a blade database that can serve as an aid technological analysis of blade for future documentation, research discoveries throughout the Coastal Plain endeavors and comparative studies. and Piedmont of South Carolina. A Moreover, a template form for recording sample of 21 blades from isolated specific criteria important for classifying contexts were examined and compared to blades is also provided and can increase the attributes of blades from the Topper public awareness of these artifact forms.

114 Clovis Blade Technology

Blade Technology United States. As such, the possibility for variation to exist in blade technologies A blade is a specialized, elongated across and within different geographical form of flake, detached from a prepared regions remained largely unknown based core. Blades are useful blanks for a upon the limited number of assemblages variety of cutting and scraping activities that have been examined. Apart from the and are efficient uses of raw material in Southern Plains and Southwest, terms of total length of cutting edge from a additional studies have documented the given mass of stone (Collins 1999; discovery of Clovis blades from other Whitaker 1994). The presence of blades regions of the U.S., most notably the from Clovis aged deposits has been Southeast. In this region, blades have known since the 1962 description of a been reported from habitation sites, cache of 17 blades recovered from the quarries, surface contexts, and from lithic Blackwater Locality No. 1 by F.E. Green manufacture sites where such artifacts (Green 1963). Since this discovery, were produced (Broster and Norton 2009; blades have been recovered from Clovis Ellerbusch 2004; Sain 2010; Steffy and sites in many regions throughout North Goodyear 2006). In the Central Savannah America, with the majority reported from River Valley of South Carolina, blades Pleistocene kill sites and caches in the have been recovered alongside fluted U.S. Southwest and Southern Plains projectile points from excavated contexts (Kilby 2010, 2011, 2013). at the Topper and Big Pine Tree sites Contemporary Clovis blade research (Sain 2010a, 2012; Sain and Goodyear has focused on the identification of 2012; Smallwood et al. 2013). Figure 1 specific technological blade attributes as a presents a sample of prismatic blades means of distinguishing Clovis-aged recovered from stratigraphically intact blades from those representative of other Paleoindian deposits at these sites. cultures in the absence of independent Technological analyses of blades from dating control (Meltzer and Cooper 2006: regional quarry sites resulted in the 127). In a seminal report on Clovis blade discovery of cross-regional variation in technology, Collins examined blades from some attributes of these forms when the South Central U.S and found that they compared with blades from the Mid-South typically exceed 100mm in length, are and Southern Plains (Sain 2010a, 2010b). curved as opposed to straight in profile, There is also increasing evidence for exhibit small striking platforms, and have blades found among private collections flat or no bulbs of force (Collins 1999). and from isolated surface localities Consequently, by isolating these specific throughout the lower Southeast. While attributes, blades can be as temporally Clovis and Late Paleoindian points from diagnostic as Clovis points and thus can such contexts have been extensively serve as another index to Clovis reported, and distributional studies of settlement behavior; important given that these artifacts have been used to test blades are often recovered from surface models of prehistoric mobility and assemblages and from disturbed settlement subsistence systems contexts. (Goodyear 2009; Miller and Smallwood A majority of the blades examined by 2009; Smallwood 2010), less research Collins were recovered from assemblages has considered the role that blades might located in the Plains and South Central have served across similar landscapes.

115 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 1. Prismatic blades from the Big Pine Tree (top row) and Topper (bottom row) Sites (38AL23). One problem with using blades from number of significant questions isolated contexts to test settlement concerning the organization of Clovis lithic subsistence models is being able to technology. The goals of this project are: differentiate the technological attributes of (1) to determine what role, if any raw Clovis blades from those that could have material accessibility, form, and type have been produced during different time on the distribution blades; (2) to establish periods. Although blades were produced the geographical range of isolated blade during the Woodland and Mississippian discoveries relative to the spatial periods (Parry 1994), research has found distribution of lithic quarries in the that a “true blade technology was not Savannah River Valley; (3) assess used by Early Archaic peoples from the whether the geographic distribution of region and that potential blades from this blades coincides with that of fluted Clovis time period typically take the form of points of the same tool-stone; and (4) blade-like flakes (Carr and Bradbury determine whether blades from isolated 2012:90). The documentation of isolated contexts compare technologically and blade discoveries can inform about a morphologically to those recovered from

116 Clovis Blade Technology

FIGURE 2. Blade scar patterning and cross section classes at left: (A) Triangular, (B) Trapezoidal, (C) Lenticular. At right, types of modification found on blades: (A) lateral margin retouch, (B) distal retouch (Image by Darby Erd). quarry sites, and if not, how they were they are manufactured to have straight being used away from such quarries? parallel margins which are reliable for maximizing the length of cutting surface Blade Technology and Clovis per working edge, blades and blade tools Settlement Subsistence serve an advantage over other flakes produced from amorphous cores. If they Blade manufacture consists of a become dull through use, simple specific design strategy, the product of modification through chipping or retouch which is intended to yield a specialized of the blade edge allows additional use type of lithic flake resulting from the life for blade tools. reduction of a prepared core. Blades Blades designed using strategies that typically exhibit a suite of technological allow for modification would have been and morphological attributes. In general, particularly essential to humans in areas blades are defined as having two or more of the landscape where raw material was parallel removal scars on the exterior scarce. In such areas, the presence of surface, and cross sections that are blade modification indicates a high utility triangular or trapezoidal when viewed in given for these tool forms and the greater profile. The parallel scar patterns present occurrence of multi-purpose blade tools on blades reflect the systematic, removals recovered in areas of raw material of prior detachments, as opposed to scarcity implies that a premium was opportunistic reduction that may result in placed on high quality tool-stone. multi directional removals (Figure 2). Therefore, the concept of blade Morphologically, blades are at least modification can be employed to develop twice as long as they are wide. They may test implications regarding prehistoric be curved or straight in profile dependent settlement subsistence systems. If blade upon raw material form and force modification was conducted in response application during detachment. Because to raw material scarcity, then it follows

117 Tennessee Archaeology 8(1-2) Summer 2016

that blades should exhibit greater attributes of utility in the form of modification the further one travels from sources of high quality tool-stone (Goodyear 1979). Such attributes would hold great value for a specific type of blade tool that could be maintained in the field and repurposed if necessary for other tool functions.

Methods

In an effort to explore the role of blades in the organization of Clovis lithic technology, we compared the morphological and technological attributes FIGURE 3. Selected morphological of blades recovered from two known measurements of blades. Top, blades size; at bottom, index of curvature. quarry sites in the Savannah River Valley to the attributes identified on a sample of measurements were recorded with the aid 21 blades recovered from isolated surface of metric calipers. The morphological contexts throughout the Coastal Plain and attributes considered for this analysis Piedmont of South Carolina. This region include maximum blade length, maximum was chosen, and is particularly well suited width, weight, platform angle, and index of for such a study as no primary sources of curvature. Index of curvature is a ratio of high quality tool-stone are found outside total blade length divided by the distance the Central Savannah River Valley. For at the arc of greatest curvature. Once this analysis, site assemblage data was obtained, this value is in turn multiplied by taken from a sample of 333 blades from 100 to producing an “index” of curvature. the Topper site and 472 blades from the Blade curvature can enlighten on a Big Pine Tree site (Sain and Goodyear number of issues, including reduction 2012). Both sites have previously been stage, manufacture technique, tool identified as prehistoric quarry and function, and raw material morphology. habitation sites with stratified Paleoindian The platform angle of a blade is the deposits (Goodyear 1999; Smallwood angle between the platform remnant and 2010; Waters et al. 2009). All blades the longitudinal axis of the blade exterior. examined for this analysis were taken An inclinometer was used to obtain the from stratigraphically discrete, intact platform angle. Blades detached from deposits where evidence of Clovis biface conical and wedge shaped cores typically technology was also noted. have greater platform angles (but less than 90 degrees) than flakes and blade Morphological Attribute Analysis like flakes struck from bifaces (Collins 1999). This is a direct result of the angle All blades were examined, recording of applied force taken when detaching a specific morphological and technological blade from a core. In contrast, flakes attributes of the exterior and interior struck from bifaces typically have acute surfaces (Figure 3). All dimensional striking platform angles.

118 Clovis Blade Technology

Technological Attribute Analysis removal scars reflect the systematic removal of prior blades from a core. Like Eight technological attribute conditions directionality, cross section is an attribute were considered for this analysis. that may be used as an indicator of Condition categories for each reduction. Later stages of blade technological attribute are presented in production often exhibit cross sections Table 1. Platform condition and type refer that are trapezoidal in form, whereas to the condition of the blade where force triangular cross sections reflect earlier was applied during detachment. stages in the manufacture process. The Preparation of the striking platform by way form of the blade margin is one indicator of grinding, or through the removal of of how a blade detached from the core small flakes allows for added control when face during force initiation. Parallel detaching a blade from a core. The bulb margins are thought to be best desired, of force, found on the interior surface of and enable for maximum length per the blade at the proximal end can reflect cutting edge. Irregular blade margins may the technique(s) employed in blade reflect human error, inexperience, or the manufacture. Diffuse bulbs of force are use of material of lesser quality. They also flat or expanding, while salient bulbs are are a frequent by-product of initial core prominent (Collins 1999). Likewise, the reduction episodes, conducted to form by which the distal end of the blade establish ridges for subsequent blade terminates may also indicate the method detachment and to rejuvenate the core or amount of applied force, or the platform and face. presence of impurities within raw material. In addition to the attribute analysis, In most cases, feather terminations are each blade was subsequently examined most desired, whereas hinge and step for the presence of modification and terminations reflect human error, improper technological retouch. This examination applied force, or less homogenous raw was conducted with the aid of a hand held material. lens. Modification includes utilization or The number and direction of exterior retouch, and applies to any type of surface scars can inform on reduction trimming (unifacial or bifacial), at any stage, or the point in the manufacture angle, that is restricted to any margin or sequence from which a given blade was edge of an artifact (White et al. 1963). detached. Uni-directional or bi-directional Modification categories considered for the

Table 1. Technological attribute conditions for blades.

119 Tennessee Archaeology 8(1-2) Summer 2016

blade analysis include lateral retouch discovered from isolated surface contexts along one or more margins or distal from the study area. Both blades and retouch (Figure 2). Moreover, the fluted points fall within the geographic presence or absence of specific tool forms footprint of the Savannah River Valley. A on blades such as gravers or burins were total of 21 of the 25 blades were also noted. In addition to the retouch analyzed. The morphological and categories, all blades were examined for technological attributes of a sample of the occurrence of snap fractures, polish, these blades are presented in Tables 2 and natural chipping. Finally, properties of and 3, and images are illustrated in Figure raw material type, quality, and condition 5. Technological attributes consistently were recorded for each blade which can found on modified blades include multiple inform about the extent of tool utility parallel uni-directional scars of previous relative to the distance to raw material blade detachments on the exterior source. surface, cross sections that are triangular to trapezoidal in form and platform angles Results of Analysis that are typically greater than 60 degrees. All but a single blade was found to have a Figure 4 presents the distribution of bulb of force that was diffuse. Likewise, fluted projectile points and blades blade margins are predominantly parallel,

FIGURE 4. Distribution of fluted Clovis points and modified prismatic blades from Allendale Briar Creek complex. Twenty-one blades from this sample were examined for the present study (adapted from Goodyear 2015).

120 Clovis Blade Technology

Table 2. Morphological and Retouch Attributes of Blades Recovered from Isolated Surface Contexts from the Coastal Plain and Piedmont of South Carolina and Georgia.

Table 3. Isolated Blade Discoveries from Private Collections in South Carolina and Georgia.

and only one blade was found to have a maximum of 124.88 mm, with a mean profile that was excessively curved. length of 96.9 mm. Blade widths ranged Platform remnants, when present, are from 19.5 mm to 47.26 mm in width, with plain, with a single example showing a mean of 36.66 mm, and blade weights evidence of having been ground. ranged from 14 g to 68.3 g with a mean of Morphological attributes were 30.8 g. When lateral edge angles on recorded for 13 of the 21 blades from the these blades were considered, the results study sample. Of these, seven blades are show that they frequently exhibit angles complete. The remaining blades include that are acute, and range from 30-45 five proximal fragments and a single degrees. medial segment. In terms of morphology, The sample of blades from isolated complete blades were found to range in surface contexts was examined for the length from a minimum of 74.38 mm to a presence or absence of modification. The

121 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 5. Modified blades recovered from isolated, non-quarry related contexts in South Carolina and Georgia. results of this analysis identified 16 (76%) scarcity implies that a premium was likely blades that exhibit modification. placed on high quality tool-stone in this Modification typically consists of unifacial region. Moreover, the results demonstrate retouch along one or both margins. Four that lithic tools that exhibit the greatest blades exhibit retouch along both margins attributes of utility (modification) are more and an end, while two additional blades often recovered at greater distances from have retouch only along the margins. At sources of high quality tools-tone. In other least one blade has been modified to words, there is a trend in greater tool create a multi-functional tool, including a utilization with distance from raw material graver spur at the distal terminus. The source. presence of blades that may serve as multi-functional tools indicates a high Comparative Study utility given for these tool forms. The higher quantity of multi-purpose blades, The sample of blades recovered from recovered in areas of raw material isolated surface contexts was compared

122 Clovis Blade Technology

to the attributes of blades recovered from Topper or Big Pine Tree sites. The results the Topper and Big Pine Tree sites, two of the t-test confirm this finding for artifact known quarry related sites in the Central length, and demonstrate that off-site Savannah River Valley (Figure 6). For this complete blades are statistically longer analysis, site assemblage data was taken than complete blades from Big Pine Tree from a sample of 333 blades from the or from Topper. However, when artifact Topper site, including 114 complete width and weight were considered, the blades, and 472 blades from the Big Pine blades from the isolated contexts only Tree site of which 314 are complete (Sain exhibit a statistical difference when 2012; Sain and Goodyear 2012; see compared to the blades from Big Pine Table 4). Both sites have previously been Tree. Although larger, blades from off-site identified as prehistoric quarry/ isolated contexts are not statistically wider manufacture and habitation sites in or heavier that blades from Topper at the Allendale, County South Carolina. .05 significance level. The same patterns Furthermore, both sites also have were found when blade weights were stratified Paleoindian deposits. examined. T-tests were conducted in order to determine if there exists any statistical difference in the morphological attributes of blades from each sample and the results of these tests are presented in Tables 5- 7. For this analysis, probability values of less than .05 are considered statistically signifi- cant. Attributes examined for the analysis include maximum blade length, maximum blade width, and weight. In addition to the t- tests, artifact plot maps were created that display the distribution of each assemblage by the attributes of length and width (Figures 7-9). This analysis allows for a visual comparison of each blade assemblage with the FIGURE 6. Topographic map showing location of the specific objective to note the Topper and Big Pine Tree sites (Allendale County, South presence, absence or degree of Carolina). morphological attribute variation across space. Table 4. Mean Morphological Attributes of Blades from The results of the Isolated Contexts Compared to the Attributes of Blades comparative analysis show that from Known Quarry Sites in the Savannah River valley. the blades recovered from isolated contexts are typically longer, wider, and heavier than blades recovered from either the

123 Tennessee Archaeology 8(1-2) Summer 2016

Table 5. Results of Students T-Test Comparing the Length of Blades from Isolated Contexts to the Length of Blades from Known Clovis Quarry Sites in the Savannah River Valley.

Table 6. Results of Students T-Test Comparing the Width of Blades from Isolated Contexts to the Width of Blades from Known Clovis Quarry Sites in the Savannah River Valley.

Table 7. Results of Students T-Test Comparing the Weight of Blades from Isolated Contexts to the Weight of Blades from Known Clovis Quarry Sites in the Savannah River Valley.

FIGURE 7. Comparison of maximum blade length versus width (mm) for Topper Clovis blades (blue) and those recovered from offsite contexts (red). Colored lines represent the mean length for each sample.

FIGURE 8. Comparison of maximum blade length versus width (mm) for Big Pine Tree Clovis blades (blue) and blades recovered from offsite isolated contexts (red). Colored lines represent the mean length for each sample.

124 Clovis Blade Technology

FIGURE 9. Comparison of maximum blade length versus width (mm) for offsite modified blades (yellow), Topper modified blades (blue), and Big Pine Tree Modified blades (red). Colored lines represent the mean lengths for blades from each sample.

form. Modification frequently consists of retouch along one or more exterior margins of the blade. By contrast, only three percent (3%) of blades from Topper, and fifteen percent (15%) of blades from Big Pine Tree are modified. The high percentage of modified blades found afield is indication that once used and removed from the local quarry, blades were being rejuvenated in regions of limited high quality raw material. Moreover, a number of blades recovered from isolated off-site contexts appear to have been modified into multifunctional tools. Such blades are rare at Topper and Big Pine Tree, where only a single example has been identified to date. Of the off-site sample, the blades that exhibit the greatest extent of modification are typically shorter in maximum length, exhibit higher proportions of tools per working edge, and are also found at greater distances from potential sources of origin than are unmodified blades. FIGURE 10. Modified blades from the Big The modified blades from Topper, Big Pine Tree Site (top) and Topper Site (bottom). Pine Tree, and isolated surface contexts In an effort to assess blade utility, I were examined by morphology. The examined each blade from the off-site results of a T-test show no statistically isolated sample noting the presence and significant difference was in blade length degree of modification. The results were between the off-site isolated sample and then compared to the sample of modified the modified blades from Topper (Table blades from the Topper and Big Pine 8). It would appear that longer, wider sample (Figure 10). The results of this blades produced at quarries were those analysis show that most blades from off- best suited for use, and were the blades site contexts exhibit modification in some most frequently carried away from for

125 Tennessee Archaeology 8(1-2) Summer 2016

Table 8. Results of Students T-Test Comparing the Attributes of Modified Blades from Topper to Blades from Off-site Contexts (modified blades are rare at Topper, comprising only 3% of the blade assemblage).

uses afield. Many blades recovered from from these materials are not present quarry sites may have been failures or within the study sample. Apart from rejects of the manufacture process. Allendale Coastal Plain chert, Moreover, as quarries are areas for raw metavolcanic materials originating from material extraction and stone tool North Carolina are the most widely production, it would make sense for a exploited tool-stones and the regional wide range of blade forms to have been occurrence of Clovis projectile points discarded at or near the quarry. Thus, produced of these materials lends support blades that fit a specific criterion, in this to the conclusion that chert was not the case, those that were long and wide sole material exploited by Paleoindian enough to serve a desired function, were inhabitants of the region (Daniel and the pieces most likely to have been Goodyear 2006, 2013; Goodyear 2010;). selected for transport and to have been However, that blades were not produced used away from the quarry. If in fact this is of lesser quality tool-stone, even when the case, then blade utility (extent of such materials were locally available, is modification) should increase with evidence for selection of high quality raw distance from source. material which allowed maintenance of a All blades from the study sample were highly curated technology. It is also examined by raw material type. Based on possible that only cryptocrystalline raw the geographical distribution of isolated materials of high quality and of large blade discoveries presented in Figure 5, package sizes allowed for the production all but one example is a product of of blades that met specific size thresholds Allendale Coastal Plain Chert. The near needed to carry out required tasks. exclusive use of Allendale Coastal Plain Chert for the manufacture of blades Development of a Regional Blade across the region lends credence to the Database notion that tool-stone was a significant factor in Paleoindian settlement mobility One byproduct of this study is the systems (Goodyear et al. 2009, Goodyear development of a regional blade database and Steffy 2003). Other lithic raw material that can serve as an aid to future artifact sources such as quartz, quartz crystal, documentation, and comparative and quartzite are locally available in the analyses. A data-entry form for recording Piedmont and Coastal Plain and specific criteria important for classifying Piedmont of South Carolina and Georgia blades has been developed. This form (Novick 1978). Moreover, various highly can serve as a data entry sheet that can fossiliferous cherts such as Black Mingo subsequently be used to increase public and Wyboo are present in the Santee awareness about blades and blade tools, River Valley, however blades produced especially since they can often be found

126 Clovis Blade Technology in private artifact collections. The data Clovis lithic technology in the Southeast entry form is divided into two sections, U.S. including: (1) primary background information; and (2) observable artifact Conclusions attributes. The primary background section is necessary for recording A preliminary analysis of artifact information such as the location and date attributes from a small sample of blades of discovery, the owner/repository of the from isolated surface contexts recovered artifact, as well as any initial visual artifact from the Coastal Plain and Piedmont of observations. Such observations may South Carolina and Georgia demonstrate include raw material type or condition, the that patterns of variation exist in blade presence or absence of exterior surface morphology and tool utility at a spatial cortex, or the stage in the reduction scale and that raw material accessibility process as represented by the number of does have some role on the distribution of dorsal scars, and the presence and extent blades across the study area. Specifically, of modification. the results of this study show that Section two of the data entry form although blades from isolated contexts provides space to document the compare technologically with those morphological and technological identified from quarry sites, blade conditions of each blade. Space morphology and to an extent tool utility necessary for the documentation of all tend to vary in accordance with distance morphological attributes is allotted on the to raw material source locations. left column of the data entry form, with an Clovis inhabitants of South Carolina area for the technological attribute and Georgia were dependent upon high conditions provided on the right. If quality tool-stone for the manufacture of available, a photograph or sketch of the blades. The results show that blades blade may be attached on page two of the produced from Allendale chert are found recording form. at distances of up to 70km from the The blade recording form presents an nearest known raw material sources, thus initial step at formulating a standardized establishing a geographical range for method to identify and document Clovis isolated blade discoveries relative to the blades when encountered in the field. The spatial distribution of known lithic quarries results of this project provide an example in the Savannah River Valley. Moreover, of the usefulness of a comparative blade the geographic distribution of blades database that can be used to inform about coincides with the distribution of fluted the organization of Clovis lithic technology Clovis points of the same tool-stone. In in the region. Specifically, this can be areas of limited raw material access, or accomplished through the proper where high quality sources were scarce, recording of specific technological and tool-stone was conserved, resulting in morphological attributes found on blades. longer life-spans for blade tools, as well The most important objective of this as increased tool utility. Based on these project is to provide the resources findings, the known geographical necessary to develop a regional database distribution of blades across the lower which may serve to entice public interest Southeast suggests that these artifact in blades, and ultimately to broaden our forms were integral to Clovis adaptation knowledge regarding the organization of and technological organization.

127 Tennessee Archaeology 8(1-2) Summer 2016

Given the small number of blades expanded Paleoindian projectile point surveys, identified from isolated localities to date, provided new insight regarding Clovis blade technology, and has made available to the general one could conclude that blades were not public data from these projects, as part of the integral to Clovis adaptations in the study Paleoindian Database of the Americas (PIDBA). area. However, research has shown that The present study on blade technology in the continued efforts to systematically Southeast U.S. was heavily influenced by prior assemble and share data on Paleoindian research efforts such as those conducted by John Broster, and without efforts such as these, the final artifacts have generated new insight product of this study would not have been about patterns of land use, demographic possible. trends and raw material utilization for the Paleoindian period (Anderson et al. 2010). References The compilation of locational and attribute information on over 30,000 fluted points Anderson, David G., D. Shane Miller, over the past two decades is one example Stephen J. Yerka, J. Christopher of the value of Paleoindian data recording Gillam, Erik N. Johanson, Derek T. projects, and demonstrates that a Anderson, Albert C. Goodyear, and substantial amount of data can be Ashley M. Smallwood. acquired within a relatively short amount 2010 PIDBA (Paleoindian Database of of time. As such, it is hoped that the the Americas) 2010: Current Status development of a regional blade database and Findings. Archaeology of may generate interest in blades and Eastern North America 38:63–90. ultimately lead to the identification of additional blades that may reveal Bradley, Bruce, Michael Collins, and important information about Clovis Andrew Hemmings technological organization in the region. 2010 Clovis Technology. International Although this project presents a Monographs in Prehistory, number of intriguing findings, it is Archaeological Series 17. Ann important to note that the results should Arbor. be considered as preliminary, as the study sample used for the comparative analysis Broster, John B., and Mark R. Norton was relatively small (n=21). Additional 2009 The Sinclair Site (40WY111): A data are therefore needed to confirm or Clovis quarry along the Buffalo refute the results of this analysis. Through River in Wayne County, public outreach and assistance in the Tennessee. Current Research in documentation and proper recording of the Pleistocene 26:35–36. blades, it is hoped that this research serves to impart valuable insight and Collins, Michael knowledge about the preservation and 1999 Clovis Blade Technology. Texas accurate dissemination of cultural Archaeology and Ethnohistory heritage. Series. The University of Texas Press, Austin. Acknowledgements. John Broster has contributed to our knowledge and understanding of Daniel, I. Randolph and Albert C. Southeastern Archaeology over the course of his career, and in this capacity has been an inspiration Goodyear to the work of numerous individuals. Research 2006 An Update on the North Carolina conducted by Broster has led to the identification Fluted Point Survey. Current of new Paleoindian sites across the Southeast,

128 Clovis Blade Technology

Research in the Pleistocene 23:88- Implications for Paleoindian 90. Mobility Patterns and Exchange. 2013 Clovis Macrobands in the South Carolina Antiquities 42:40- Carolinas. Poster presented at the 41. Paleoamerican Odyssey Conference, Santa Fe, New Goodyear, Albert C., Keith Derting, D. Mexico. Shane Miller, and Ashley M. Smallwood Ellerbusch, Elijah C. 2009 Exotic Clovis Stone Tools from the 2004 Paleoamerican Prismatic Blade Topper Site, 38AL23, Allendale Economy from the Nuckolls Site County, South Carolina. Current (40HS60), Lower Tennessee River Research in the Pleistocene 26:60- Valley, Tennessee. Current 62 Research in the Pleistocene 21:35- 38. Goodyear, Albert C., III and Kenn Steffy 2003 Evidence of a Clovis Occupation at Goodyear, Albert C. the Topper Site, 38AL23, Allendale 1979 A Hypothesis for the Use of County, South Carolina. Current Cryptocrystalline Raw Materials Research in the Pleistocene 20:23- among Paleoindian Groups of 25. North America. Research Manuscript Series 156, South Green, Frank E. Carolina Institute of Archaeology 1963 The Clovis blades: an important and Anthropology, University of addition to the Llano complex. South Carolina, Columbia. American Antiquity 29:145–165. 1999 The Early Holocene Occupation of the Southeastern United States: A Kilby, J. David Geoarchaeological Summary. In 2010 Toward Determining the Functions Ice Peoples of North America, of Clovis Caches. Paper presented edited by Robson Bonnichsen and at the 75th Annual Meeting of the Karen L. Turmire, pp. 432-481. Society for American Archaeology. Oregon State University Press. St. Louis, Missouri. Corvallis. 2011 A Regional Perspective on Clovis 2006 Recognizing the Redstone Fluted Blades and Clovis Blade Caching. Point in the South Carolina Paper presented at the 76th Paleoindian Point Database. Annual Meeting of the Society for Current Research in the American Archaeology, Pleistocene 23:100-103. Sacramento, California 2009 The Allendale-Brier Creek (ABC) 2013 Clovis Caches and the Early Clovis Complex of the Central Paleoindian Record of North Savannah River Valley. Paper America. In Peuplements et presented at the 74th Annual Préhistoire de l'Amérique, edited by Meeting of the Society for Denis Vialous. Muséum National American Archaeology. Atlanta. d'Histoire Naturelle, Paris, France. 2010 Lithic Raw Material Studies in South Carolina and Their

129 Tennessee Archaeology 8(1-2) Summer 2016

Meltzer, David J., and Judith R. Cooper Contemporary Lithic Analysis in the 2005 On Morphometric Differentiation of Southeast, Problem, Solutions, and Clovis and Non-Clovis Blades. Interpretations, edited by Philip J. Current Research in the Carr, Andrew P. Bradbury, and Pleistocene 23:143–145. Sarah E. Price, pp. 42-54. University of Alabama Press, Miller, D. Shane and Ashley M. Tuscaloosa. Smallwood 2009 Beyond Stages: Modeling Clovis Smallwood, Ashley M. Biface Production at the Topper 2010 Clovis Biface Technology at the Site (38AL23), South Carolina. Topper Site, South Carolina: Paper presented at the 66th Evidence for Variation and Annual Southeast Archaeology Technological Flexibility. Journal of Conference. Archaeological Science 37:2413– 2425. Novick, Lee 1978 Prehistoric Lithic Material Sources Smallwood, Ashley M., D. Shane Miller, and Types in South Carolina: A and Douglas Sain Preliminary Statement. South 2013 Topper Site, South Carolina: An Carolina Antiquities 10:422-437. Overview of the Clovis Lithic Assemblage from the Topper Sain, Douglas A Hillside. In In the Eastern Fluted 2010a A Technological Analysis of Clovis Point Tradition, edited by Joseph Blades from the Topper Site, A. M. Gingerich, pp. 280-298. 38AL23, Allendale County, South University of Utah Press. Carolina. Current Research in the Pleistocene 27:136-139. Steffy, Ken and Albert C. Goodyear 2010b Clovis Blade Technology at the 2006 Clovis Macro Blades from the Site. Unpublished MA thesis, Topper Site (38AL23), Allendale Eastern New Mexico University. County South Carolina. Current 2012 Clovis Blade Technology at the Research in the Pleistocene 26: Topper Site (38AL23), Assessing 147–149. Lithic Attribute Variation and Regional Patterns of Technological Waters, Michael R., Steven L. Forman, Organization. Occasional Papers Thomas W. Stafford and John E. No. 2, Southeastern Foss. Paleoamerican Survey, South 2009 Geoarchaeological Investigations Carolina Institute of Archaeology at the Topper and Big Pine Tree and Anthropology, University of Sites, Allendale County, South South Carolina. Carolina. Journal of Archaeological Science 36:1300-1311. Sain, Douglas A. and Albert C. Goodyear 2012 A Comparison of Clovis Blade White, Anta M, Lewis R. Binford, and Technologies at the Topper and Mark L. Papworth Big Pine Tree Sites, Allendale 1963 Miscellaneous Studies in Typology County, South Carolina. In and Classification. Museum of

130 Clovis Blade Technology

Anthropology, University of Michigan Anthropological Papers No. 19.

Whittaker, John C. 1994 Flintknapping Making and Understanding Stone Tools. The University of Texas, Austin.

Douglas Sain Terracon Consultants, Columbia, South Carolina 29208

Albert C. Goodyear South Carolina Institute of Archaeology and Anthropology University of South Carolina Columbia, South Carolina 29208

131 MAKING A DIFFERENCE: JOHN B. BROSTER AND PALEOINDIAN ARCHAEOLOGY IN TENNESSEE

David G. Anderson

Every state needs someone like John Mississippian phases, as well as if he Broster, who with younger colleagues like knew where Paleoindian sites were to be Mark Norton, took on the role of finding, found in the state. He responded to both compiling, and publishing information on questions in great detail, providing a lot of Paleoindian archaeology locally beginning primary data that helped me produce some 30 years ago. Few states have such maps for these periods that covered most a person, much less someone who has of eastern North America (Anderson done so much for so long, making John’s 1990a, 1991; John’s contributions are approach an example that others should acknowledged in each paper). That adopt, and whose legacy will hopefully marked the beginning of a long friendship continue in Tennessee now that he is and many subsequent exchanges of formally retired. Fortunately, John is still information, something he has always very much in the game, as he would say, generously done when I or others have if moving a little slower at it. The asked, and what makes John Broster organizers and contributors deserve our such a great colleague. He will give you thanks for producing this series of essays the shirt off his back, or more accurately, in honor of John Broster, one of our access to the primary data he has region’s leading Paleoindian researchers collected, if he thinks it will advance and personalities, and much more archaeology. Not every archaeologist, besides. Indeed, as Shane Miller and especially in the Paleoindian research Jesse Tune note in their introduction, and community—sometimes as rowdy and as Mike Moore, Kevin Smith, Aaron Deter- cantankerous a group as there ever has Wolf, and David Stuart summarize so been in American archaeology—is as capably in their biographical overview, open with their data and ideas and as John is not just a Paleoindian collegial in helping their colleagues as archaeologist, but has worked on sites of John Broster has been down through the all periods in Tennessee, particularly on years. sites of the Mississippian and historic It quickly became clear that John periods with his colleagues at the Division Broster was the person most of Archaeology and beyond. But he also knowledgeable about Paleoindian conducted fieldwork in the Southwest, on archaeology in Tennessee, and for that the southern Plains, in the Netherlands, reason I invited both him and Mark Norton and particularly in Mesoamerica, where to come to a workshop on “Paleoindian his legacy in and out of the field remains and Early Archaic Period Research in the the subject of legend. Southeast” that was held in Columbia, I first began interacting with John South Carolina in September 1991. That Broster in the late 1980s, when I asked workshop became the basis for the book him about prehistoric site distributions in The Paleoindian and Early Archaic central Tennessee, notably the Southeast (Anderson and Sassaman occurrence of Late Woodland and 1996), which has a chapter that John and

132 Making a Difference

Mark Norton wrote summarizing Tennessee archaeology. No stunt double Tennessee Paleoindian archaeology, was ever necessary… as they document, documenting the impressive record of John has done an impressive amount of sites and artifacts that have been found in field archaeology, where he did all his the state (Broster and Norton 1992, own stunts and a lot of hard work besides, 1996). John’s many papers, in fact, and he remains a larger than life figure to constitute a substantial portion of the all familiar with him. The words Moore and published record on modern Paleoindian his colleagues used to describe John, that archaeology in Tennessee. But he didn’t are echoed in the other papers herein, are do the work by himself, as he has always respect, trust, and admiration. Indeed, been quick to state… John has always John’s work with the avocational been generous with sharing authorship community in Tennessee is a primary and credit, another mark of a true scholar. reason why we know as much as we do As an aside, I video-filmed the 1991 about the Paleoindian archaeological conference, and John’s talk is crisp and record of the state. His career reminds us lucid, recounting the numbers of through example of the importance of Paleoindian projectile points by type working with avocationals, reaching out to recorded as of that time in the survey the public, and building relationships with started in 1988, as well as information on private citizens and professional the Paleoindian sites like Johnson he was colleagues alike based on mutual of trust, working on at the time. Interestingly, in the respect, and generosity. film John’s hair is dark black, and he Fieldwork has also been a major part didn’t have quite the pronounced drawl of John’s life, and a lot of our knowledge nor the somewhat grayer hair he has of the archaeological record in today, both of which may be the result of Tennessee, particularly for the early subsequent decades working with periods, is largely based on his work, and Tennessee landowners, collectors, and that of his colleagues and collaborators state agency officials. When I asked John like Mark Norton, Emanuel Breitburg, and at the 1991 meeting how he pronounced many others. He has participated in the his last name so we could introduce him, discovery and documentation of such he said “Brewster, rhymes with rooster!” remarkable Paleoindian sites as Coats- This is very appropriate… and not for the Hines, Carson-Conn-Short, Johnson, and reasons some may be thinking, but Widemeier, among others. There is a because like the gamecocks of South saying in archaeology that luck is an Carolina, John has always been a important part of our research, at least in scrapper, scratching and pecking river the discovery phase, but you have to be banks, plowed fields, and private good at fieldwork and research too. John collections in search of early sites and has been both lucky at finding sites and assemblages. good at wresting important information Michael Moore, Kevin Smith, Aaron from them, making his own luck through Deter-Wolf, and David Stuart’s paper hard work throughout his career. Indeed, provides an excellent overview and tribute my respect for John was heightened even to John’s life and professional career, more than I thought possible when he helping younger readers understand why came out day after day to the Bells Bend John is such a presence in both the field project Shane Miller, Thad Bissett, technical and personal sides of modern Stephen Carmody and I were running

133 Tennessee Archaeology 8(1-2) Summer 2016

FIGURE 1. Shane Miller and John Broster at the Bells Bend Archaeological Project area near Nashville, August 2012. The Widemeier site is in the background, which probably explains the golden rays of light shining on them.

FIGURE 2. John Broster, Aaron Deter-Wolf, and Shane Miller discussing strategy, with Sarah Walters in the background, during the Bells Bend Archaeological Project near Nashville, August 2012.

along the Cumberland River near Mark Norton started in 1988, was one of Nashville in 2010 and 2012, looking for many people whose archaeological career early sites (Figures 1–3) (Miller et al. was strongly and positively influenced by 2012). John visited and worked with us in John Broster. Indeed, one of John’s the often 100 plus degree summer heat, lasting contributions to the field is the providing insight into what we were influence he has had on young people, finding, and showing us locations along turning them from interested students to and near the river he had visited where solid scholars. John does this through further research might be profitable. He example, with his infectious interest and also, I might add, through his method of enthusiasm for all things Paleoindian, instruction using colorful oral historical delivering information with a charm and accounts, left a lasting impression on the confident expertise that epitomizes a true undergraduate and graduate students on southern gentleman, albeit perhaps with a the project. touch of the rascal or gunfighter thrown in. Jesse Tune, whose paper herein As Tune recounts, nearly 5,500 provides an update on the Tennessee Paleoindian points have been Fluted Point survey John Broster and documented in Tennessee to date by

134 Making a Difference

FIGURE 3. Tom Pertierra, John Broster, Shane Miller, Aaron Deter-Wolf, and Valerie McCormack, at a site in the Bells Bend Archaeological Project area near Nashville, August 2012. John always wore broad brimmed Mexican hats, making him easily recognizable at great distances. Broster and Norton, the largest number of attribute data on Dalton points been points with systematically recorded systematically recorded statewide, attribute data recorded for any state or although recently recording these points province in North America. This has begun in the Carolinas (Anderson et information is available online in PIDBA, al. 1990; Ledbetter et al. 2008; McGahey the Paleoindian Database of the 2004; Smallwood et al. 2015). Given the Americas, where John and his colleagues large numbers of points recorded in have been contributing their data for Tennessee, mostly from the central and decades (Anderson 1990a, 1990b; western parts of the state, distributional Anderson et al. 2010:66, 2015:16; Broster analyses must take into account the 1989; Broster and Norton 1992, 1996; impact individual researchers or teams of Broster et al. 2013). Importantly, Broster researchers can have on sample sizes and Norton record all Paleoindian points, and locations, in the state and beyond, including both fluted and unfluted forms. over the larger region. The fact that The Tennessee total accounts for over Paleoindian points in Tennessee are 40% of all the recorded Paleoindian points concentrated in the center of the state in PIDBA from the southeastern United may be due to the presence of major States, and by far the largest numbers of drainages there, like the Tennessee and recorded Clovis, Cumberland, and Dalton Cumberland Rivers, but it also likely points (Anderson et al. 2015:16). Indeed, reflects a proximity to John Broster’s in only three states in the region, office at the Tennessee Division of Tennessee, Mississippi, and Georgia, has Archaeology in Nashville, where he has

135 Tennessee Archaeology 8(1-2) Summer 2016

worked for decades. What the Tennessee point occurrence and density. But to data does show, additionally, is that we succeed at this, to bring our knowledge of need many more people across North Paleoindian land use into better America as conscientious as John and resolution, we will continue to need to find Mark in recording data on early sites and and recruit new people to document sites artifacts, and as open and helpful in and collections, as John himself has done sharing it. The Tennessee fluted point in Tennessee. survey, as Tune and others have shown Parish and Finn also make an (most notably John Broster and Mark important contribution to recognizing Norton), provides a wealth of data useful variability in lithic raw materials to the documentation of morphological outcropping within Tennessee as well as variability in Paleoindian bifaces and their on artifacts found on sites in the state. occurrence on the landscape. The authors note more problem oriented, The paper by Parish and Finn shows fine grained lithic sourcing analyses will what can be learned from the careful be needed if we are to have greater documentation of private collections, confidence in our interpretations of especially in areas where Paleoindian prehistoric settlement locally, but their settlement is suspected, but few sites or results are a good step for determining assemblages have been documented. the scale over which lithic raw materials Their work along the lower Tennessee and perhaps people moved. Whether and River, revealing a previously unsuspected where Paleoindian band-macroband major concentration of fluted point sites, settlement systems were located will shows that low density areas or voids in require analyses like that conducted by our current distribution maps may, unless Parish and Ryan over vast areas, but their demonstrated otherwise, reflect gaps in conclusion that Paleoindian sites along our data and survey coverage rather than the Lower Tennessee River may where people were or weren’t living in the represent “occasional congregation areas past, as a number of researchers have for three macro-bands centered locally suggested (e.g., Buchanan 2003; Lane along the Western Highland Rim, and Anderson 2001; Miller and Carmody, Northern Alabama and the confluence of herein; Prasciunas 2011; Shott 2002, the Ohio and Mississippi Rivers” is 2005). Loebel and colleagues have made certainly something that can be evaluated similar observations in the upper Midwest in the years to come. I don’t doubt that based on work with private collections in aggregation areas may be present, but I Wisconsin and Illinois; areas in those do question why they need to be between states long assumed to have few or at the margins of the ranges of groups Paleoindian sites and artifacts, upon based elsewhere, as both Daniel (2001) extended analysis, have proven to be and Parish and Finn suggest. I argue anything but barren (Koldehoff and Loebel instead, as I have for decades (Anderson 2009; Loebel 2005). While I don’t expect 1996a; Anderson and Hanson 1988), that the major concentrations of fluted points macroband aggregation loci were simply across the continent to change markedly resource rich areas within the ranges of in the decades to come, or be individual bands that were a convenient overshadowed by new areas, I do believe (and ideally memorable and resource rich our maps will become far more detailed, areas) place for groups from other areas providing more accurate information on to visit and be hosted by the home group,

136 Making a Difference

with the favor reciprocated at similar Archaic assemblages or indeed artifacts locations in other group ranges in of any kind are rare in the upland and subsequent years or decades, depending mountainous areas of the Midsouth, at on the interval between aggregation. least until the Dalton period, something Adopting such a perspective implies a long noted but only poorly understood, different use of the landscape than that and attributed to the area being less inferred in lithic-centric models by favorable for settlement than lower lying Gardner (1983, 1989), Goodyear (1989) larger river valleys, and hence among the and others, that have these early peoples last areas to be occupied (e.g., Anderson tethered (i.e., living at or near) to specific 1990a; Lane and Anderson 2001; Miller lithic raw material sources. I instead 2014; Walthall 1998). As Franklin and his believe that Paleoindian people in the colleagues note, this absence may be Southeast lived more widely over the more due to the history of archaeological landscape, in areas that had attractions research than where past peoples were other than or in addition to knappable actually living, since little intensive stone. They probably would have lived systematic subsurface archaeological around quarry areas only if their band survey has occurred in the highland resided in the immediate area, or if areas, and less extensive cultivation located elsewhere, when they needed to exposing surface areas where artifacts go there as part of specialized lithic might be found. Based on their discovery, procurement forays, perhaps as part of Franklin et al. argue that we need to periodic aggregations. Given reconsider the traditional assumption that ethnographic evidence that long distance Paleoindian use of this part of the forays for raw materials were not landscape may have been minimal. Miller uncommon among some hunter and Carmody’s analysis herein, using a gatherers, we should be evaluating what large sample of site file data from the other parts of the southeastern landscape general region, in fact, demonstrates how were attractive to early populations (Speth unusual Rock Creek Mortar Shelter is et al. 2013). Settlement may have given our present state of knowledge occurred anywhere resources sufficient about site occurrence, highlighting the for groups needs were present, and site’s importance, and the need to aggregation loci, in this view, could have determine the extent of bias in our survey been wherever large numbers of people coverage. The Paleoindian assemblage could be sustained for a few days or from Rock Creek Mortar shelter is, in fact, weeks, ideally at dramatic places on the extensive and includes well-made blades, landscape that could be easily which are characteristic of Clovis remembered and reached (Anderson assemblages elsewhere in North America 1990a, 1995; Miller 2011, 2014). (e.g., Collins 1999; Sain 2011; Tankersley Jay Franklin and his colleagues work 2004). Blade-like flake tools have also at the Rock Creek Mortar Shelter been found in Dalton levels at the Dust (40PT209), documenting potential Cave and Stanfield Worley sites in Paleoindian occupations on the Upper Alabama, where they were worked into Cumberland Plateau, is an important and scraping tools (DeJarnette et al. 1962; unexpected contribution to Tennessee Hollenbach 2009; Sherwood et al. 2004), archaeology. As the authors note, which fits with their association with a potential stratified Paleoindian and Early reworked Greenbrier Dalton at Rock

137 Tennessee Archaeology 8(1-2) Summer 2016

Creek Mortar Shelter. Until further work cal yr BP, corresponding to the Younger has occurred, the possibility also remains Dryas chronozone (Anderson 2001:152– that the blades from Rock Creek Mortar 156; Anderson and Sassaman 2012:5; Shelter could be from earlier occupations. see also Waters and Stafford 2013 who The evidence from the use-wear analysis argue for a similar arrangement, with their and the diversified nature of the early Early Paleoindian Pre Clovis era assemblage at Rock Creek Mortar Shelter described as the ‘Exploration’ period). clearly point to a range of activities Shane Miller and Stephen Carmody’s occurring, indicating site use, and hence paper, as noted, offers a contrast to the occupation and use of this upland area, discoveries at Rock Creek Mortar Shelter was unlikely to have been temporary or through a theoretically informed analysis limited. of site file data. In it they argue that the Finally, a comment on chronology is Upper Cumberland Plateau was not warranted, given Franklin et al.’s settled intensively until the Late observation that tying cultures to specific Paleoindian period and particularly in time periods makes intersite comparisons Early Archaic times, well after the Middle difficult. I agree, and argue that work Paleoindian Clovis period. Their paper directed to producing stratigraphically well summarizes and serves as an excellent controlled and dated assemblages, like he example of the increasingly theoretically and his colleagues are doing at Rock sophisticated and data rich research that Creek Mortar Shelter, is absolutely critical has been occurring in the Southeast in if we are to understand what was recent years examining culture change occurring at specific times over the region. over time during and immediately after the Because archaeological assemblages Paleoindian era (e.g., Anderson et al. typically overlap and even within known 2015; Gingerich 2013; Miller 2011, 2014; phases do not occur at the same time Miller and Gingerich 2013; Morrow 2014, everywhere, use of traditional stage 2015; Smallwood 2012; Smallwood et al. formulations based on the occurrence of 2015; Thulman 2006, 2009). Miller and site or artifact types can be problematic. Carmody’s analysis, like that by Finn and Sometimes diagnostic artifacts are all we Parish, shows that our assumptions about have to work with, but whenever possible Paleoindian settlement in the Midsouth we should use chronological or period are in need of qualification and finer terminology rather than cultural phases or grained study. There appear to be great stages when reporting and comparing differences in the occurrence of assemblages. In a framework for Paleoindian assemblages in different Pleistocene occupations proposed for parts of the region due, in part, to the Eastern North America discussed at some nature of the settlement systems in use length in a recent synthesis (Anderson et and the expansion of populations into less al. 2015:8–9), the Early Paleoindian favorable habitats, facilitated by changes period corresponds to assemblages in climate and vegetation. >13,250 cal yr BP., the Middle The major decrease in Early Archaic Paleoindian period to sites dating from sites noted by Miller and Carmody at the 13,250 to 12,850 cal yr BP, not confluence of the Duck and Tennessee uncoincidentally the currently accepted Rivers is puzzling since it runs against range of the Clovis culture, and the Late theoretical expectations that such a Paleoindian period from 12,850 to 11,700 readily accessible, resource-rich area

138 Making a Difference should always be densely settled. As the such work has occurred and will for many authors state, “there is still much to learn” years to come reflects the willingness of about past occupations in the region. It the principal investigators, John Broster may be that empty areas or buffer zones and Mark Norton, and Albert C. between groups, or even large areal Goodyear, to make the data from these abandonments were occurring at this sites available to interested researchers. early date, possibly the result of migration, Both recognize that archaeologists work warfare, or changes in resource structure. best when they work cooperatively, Although these kind of large scale events something essential in fieldwork, analysis, are thought to have occurred much later and writing. Such behavior serves as a in time, there are parallels elsewhere in model of openness and data sharing all the east, such as the post Dalton Early archaeologists in the region can admire, Archaic period settlement ‘collapse’ in the and is another reason why John Broster is central Mississippi Valley (Morse and so well appreciated by his colleagues. Morse 1983), or the varied incidence of Well, that and for the great stories he tells! sites throughout the Archaic period over Sites like Topper and Carson-Conn- the region (Anderson 1996b; Sassaman Short will thus occupy the attention of 2010). scholars for generations to come, as it is The Topper site along the lower clear there is much more that can be Savannah River in South Carolina has learned from them, both through produced one of the most extensive additional fieldwork and multidisciplinary Clovis archaeological assemblages in analyses. Indeed, as recent monographic secure stratigraphic context ever found in scale reporting of the Clovis materials the Southeast. It is thus fitting that two from the deeply stratified Gault site in papers presenting new information about Texas has shown (Waters et al. 2011), it Topper—by Derek Anderson and his takes the work of a great many scholars colleagues, and Doug Sain and Albert C. to adequately report assemblages from Goodyear—are included in a collection of large, complex Paleoindian sites. Derek essays dedicated to the work of John Anderson has been the field and Broster, who with his colleagues has laboratory director for excavations into the found and worked on similarly important Clovis deposits at Topper in recent years, stratified Paleoindian sites in Tennessee overseeing with Al Goodyear a lot of the like Carson-Conn-Short, Coats-Hines, important work that has been ongoing, Johnson, and Widemeier, to name just some of which is summarized herein, just some of the many sites of this period he as John Broster has guided and assisted has helped explore (Broster et al. 1991, with research on Paleoindian sites and 1996, 2006, 2013; Broster and Barker collections in Tennessee. The recent AMS 1992; Broster and Norton 1992, 1993, dating of the hillside Clovis occupation to 1996; Deter-Wolf et al. 2011). Because about 13,000 cal yr BP based on samples the Clovis biface and blade assemblages collected by Sarah Walters, importantly, is from Carson-Conn-Short and Topper are the first date to fall within the posited so extensive and in such good context, range for this culture obtained from a collections from them are seeing stratified terrestrial Clovis site in the extensive descriptive and comparative Lower Southeast (Goodyear 2013; investigation (e.g., Sain 2011; Smallwood Walters 2013, 2016; Walters et al. 2013; 2011, 2012; Stanford et al. 2006). That Waters and Stafford 2007). Above and

139 Tennessee Archaeology 8(1-2) Summer 2016

beyond the early date, the recent work at primary data readily available. Topper demonstrates the utility of refitting In closing, I am pleased to have been analyses for assessing the integrity of a part of this session honoring the life and deposits, by showing how much accomplishments, professional and horizontally and vertically originally personal, of my friend, colleague, and conjoined artifacts have moved, and for mentor in Tennessee archaeology, John reconstructing activity areas, helping Broster. His career stands as model of determine what was occurring at the site hard work and data sharing and openness (Anderson 2011; Miller 2010). Indeed, for all archaeological researchers in the Derek Anderson has a masterful ability at region. John has also been blessed with a refitting, permitting determination of where great life partner, Diane, who he has individual knappers were sitting, and the always acknowledged as his greatest kinds of activities they were conducting, discovery and strongest supporter. One skills perhaps too infrequently employed feature of this collection of papers that or appreciated. gives me great hope for the future is the Douglas Sain and Albert C. fact that almost all the authors are half the Goodyear’s paper calling for the age of our honoree and this discussant. systematic description and reporting of There are thus a number of young Paleoindian blade data, is the kind of work scholars following in John’s footsteps… needed if we are to more completely although perhaps for their sake and understand early occupations in the longevity, given his eventful life, hopefully Southeast. Little effort has been directed not exactly in his footsteps. John, I look to the movement of blades on the forward to sitting on backdirt piles with landscape in Paleoindian times, and Sain you in the years to come watching these and Goodyear’s analysis is important in younger folks at work. showing that these artifacts moved the same distances points did, suggesting Acknowledgements. Thanks to Shane Miller and they were used or carried together, and Jesse Tune for inviting me to be a part of the session at the Southeastern Archaeological that blade modification apparently Conference in 2013 that these papers came from, increases with distance from source and to John Broster and his colleagues at the areas. The sharing of primary data is a Division of Archaeology in Nashville for their help fundamental premise of science, since the and support through the years. There are a data our analyses are based on must be number of similarities between John Broster’s career and those of Shane Miller and Jesse Tune. available for inspection if the results are to John went from Vanderbilt to the University of New be properly replicated and evaluated. Mexico, Jesse went from MTSU to American Doug Sain has been doing his part in this University and then on to Texas A&M, while Shane regard, with images as well as detailed went from the University of Tennessee to the attribute data available for hundreds of University of Arizona. All three left the east behind for the opportunity to work with distinguished blades in his thesis monograph on Topper western archaeologists and at western sites, blades (Sain 2011). Sain most recently expanding their theoretical and methodological completed a dissertation on the Pre- toolkits and exposing them to new field areas. Like Clovis deposits at Topper that includes John, Shane has conducted work in Mexico, at the almost 2000 pages of appendices of Fin del Mundo site in Sonora, and in other parts of the world like the Caribbean, while Jesse’s career primary data taken from the 30 years of is equally peripatetic. As long as they and others fieldwork at the site (Sain 2015), again like them are carrying on his work, John Broster’s highlighting the importance of making legacy will be long remembered in Tennessee archaeology.

140 Making a Difference

References Case Study from the Savannah River Basin. American Antiquity Anderson, David G. 53:262–286. 1990a The Paleoindian Colonization of Eastern North America: A View Anderson, David G. and Kenneth E. from the Southeastern United Sassaman (editors) States. In Early Paleoindian 1996 The Paleoindian and Early Archaic Economies of Eastern North Southeast. University of Alabama America, edited by Kenneth Press, Tuscaloosa. Tankersley and Barry Isaac, pp. 163–216. Research in Economic Anderson, David G. and Kenneth E. Anthropology Supplement 5. Sassaman 1990b A North American Paleoindian 2012 Recent Developments in Projectile Point Database. Current Southeastern Archaeology: From Research in the Pleistocene 7: 67– Colonization to Complexity. Society 69. for American Archaeology Press, 1991 Examining Prehistoric Settlement Washington, D.C. Distribution in Eastern North America. Archaeology of Eastern Anderson, David G., R. Jerald Ledbetter, North America 19:1–22. and Lisa D. O'Steen 1995 Paleoindian Interaction Networks in 1990 Paleoindian Period Archaeology of the Eastern Woodlands. In Native Georgia. Georgia Archaeological American Interaction: Multiscalar Research Design Paper 6. Analyses and Interpretations in the University of Georgia Laboratory of Eastern Woodlands, edited by Archaeology Series Report 28. Michael S. Nassaney and Kenneth Athens. E. Sassaman, pp. 1–26. University of Tennessee Press, Knoxville. Anderson, David G., D. Shane Miller, 1996a Models of Paleoindian and Early Stephen J. Yerka, J. Christopher Archaic Settlement in the Gillam, Erik N. Johanson, Derek T. Southeastern United States. In The Anderson, Albert C. Goodyear, and Paleoindian and Early Archaic Ashley M. Smallwood Southeast, edited by David G. 2010 PIDBA (Paleoindian Database of Anderson and Kenneth E. the Americas) 2010: Current Status Sassaman, pp. 29–57. University and Findings. Archaeology of of Alabama Press, Tuscaloosa. Eastern North America 38:63–90. 1996b Modeling Regional Settlement in the Archaic Period Southeast. In Anderson, David G., Ashley M. Archaeology of the Mid-Holocene Smallwood, and D. Shane Miller Southeast, edited by K. E. 2015 Pleistocene Human Settlement in Sassaman and D. G. Anderson, the Southeastern United States: pp. 157–176. University of Florida Current Evidence and Future Press, Gainesville. Directions. Paleoamerica 1(1):7– 51. Anderson, David G., and Glen T. Hanson 1988 Early Archaic Occupations in the Southeastern United States: A

141 Tennessee Archaeology 8(1-2) Summer 2016

Anderson, Derek T. Broster, John B., D. P. Johnson, and Mark 2011 The 2010 4x4 Meter Unit at R. Norton Topper: Preliminary Lithic and 1991 The Johnson Site: A Dated Clovis- Spatial Analyses. Legacy Cumberland Occupation in (Newsletter of the South Carolina Tennessee. Current Research in Institute of Archaeology and the Pleistocene 8:8–10. Anthropology) 15(1):16–19. Broster, John B., Mark R. Norton, Dennis Broster, John B. J. Stanford, C. Vance Haynes, Jr., 1989 A Preliminary Survey of the Paleo- and Margaret A. Jodry. Indian Sites in Tennessee. Current 1994 Eastern Clovis Adaptations in the Research in the Pleistocene 6:29– Tennessee River Valley. Current 31. Research in the Pleistocene 11:12–14. Broster, John B., and Gary L Barker 1992 Second Report of Investigations at Broster, John B., Mark R. Norton, Dennis the Johnson Site (40DV400): The J. Stanford, C. Vance Haynes, Jr., 1991 Field Season. Tennessee and Margaret A. Jodry Anthropologist 17(2):120–130. 1996 Stratified Fluted Point Deposits in the Western Valley of Tennessee. Broster, John B., and Mark R. Norton In Proceedings of the 14th Annual 1992 Paleoindian Projectile Point and Mid-South Archaeological Site Survey in Tennessee: 1988– Conference, edited by R. Walling, 1992. In Paleoindian and Early C. Wharey, and C. Stanley, pp. 1– Archaic Period Research in the 11. Panamerican Consultants, Lower Southeast: A South Carolina Special Publications 1, Tuscaloosa, Perspective, edited by David G. Alabama. Anderson, Kenneth E. Sassaman, and Christopher Judge, pp. 263– Broster, John B., Mark R. Norton, B. 68. Columbia: Council of South Hulan, and E. Durham Carolina Professional 2006. A Preliminary Analysis of Clovis Archaeologists. through Early Archaic Components 1993 The Carson-Conn-Short Site at the Widemeier site (40DV9), (40BN190): An Extensive Clovis Davidson County, Tennessee. Habitation in Benton County, Tennessee Archaeology 2(2):120– Tennessee. Current Research in 127. the Pleistocene 10:3–5. 1996 Recent Paleoindian research in Broster, John B., Mark R. Norton, D. Tennessee. In The Paleoindian Shane Miller, Jesse W. Tune, and and Early Archaic Southeast, Jonathan D. Baker edited by David G. Anderson and 2013 Tennessee’s Paleoindian Record Kenneth E. Sassaman, pp. 288– the Cumberland and Lower 297. University of Alabama Press, Tennessee River Watersheds. In In Tuscaloosa. the Eastern Fluted Point Tradition, edited by Joseph A. M. Gingerich, 299–314. Salt Lake City: University

142 Making a Difference

of Utah Press. Holocene (circa 9200 to 6800 B.C.).” In Paleoindian Research in Buchanan, Briggs Virginia: A Synthesis, edited by J. 2003 The Effects of Sample Bias on M. Wittkofski, and T. R. Reinhart, Paleoindian Fluted Point Recovery 5–51. Special Publication 19. in the United States. North Archeological Society of Virginia, American Archaeologist 24(4):311– Richmond. 338. Gingerich, Joseph A. M., editor Collins, Michael 2013 In the Eastern Fluted Point 1999 Clovis Blade Technology: A Tradition. University of Utah Press, Comparative Study of the Keven Salt Lake City. Davis Cache, Texas. University of Texas Press, Austin. Goodyear, Albert C., III. 1979 A Hypothesis for the Use of Daniel, I. Randolph Cryptocrystalline Raw Materials 2001 Stone Raw Material Availability among Paleoindian Groups of and Early Archaic Settlement in the North America. Research Southeastern United State. Manuscript Series 156, South American Antiquity 66: 237–265. Carolina Institute of Archaeology and Anthropology. Columbia: DeJarnette, David L., Edward B. Kurjack University of South Carolina. and James W. Cambron 2013. Update on the 2012–2013 Activities 1962 Excavations at the Stanfield- of the Southeastern Paleoamerican Worley Bluff Shelter. Journal of Survey. Legacy (Newsletter of the Alabama Archaeology 8(1–2):1– South Carolina Institute of 124. Archaeology and Anthropology) 17(1):10–12. Deter-Wolf, Aaron, Jesse W. Tune, and John B. Broster Hollenbach, Kandace D. 2011 Excavations and dating of Late 2009 Foraging in the Tennessee River Pleistocene and Paleoindian Valley 12,500 to 8,000 Years Ago. deposits at the Coats-Hines Site, University of Alabama Press, Williamson County, Tennessee. Tuscaloosa. Tennessee Archaeology 5(2):142– 156. Koldehoff, Brad, and Thomas J. Loebel 2009 Clovis and Dalton: Bounded and Gardner, William M. Unbounded Systems in the 1983 Stop me if you’ve heard this one Midcontinent of North America. In: before: The Flint Run Paleoindian Lithic Materials and Paleolithic Complex Revisited. Archaeology of Societies, edited by Brian Adams Eastern North America 11: 49–59. and Brooke S. Blades, pp. 270– 287. Wiley-Blackwell Publishing, Gardner, William M. New York. 1989 An Examination of Cultural Change in the Late Pleistocene and Early

143 Tennessee Archaeology 8(1-2) Summer 2016

Lane, Leon, and David G. Anderson Annual Meeting of the Society for 2000 Paleoindian Occupations of the American Archaeology, Southern Appalachians: A View Sacramento, California. from the Cumberland Plateau of 2014 From Colonization to Kentucky and Tennessee. In Domestication: A Historical Archaeology of the Appalachian Ecological Analysis of Paleoindian Highlands, edited by Lynne P. and Archaic Subsistence and Sullivan and Susan C. Prezzano, Landscape Use in Central pp. 88–102. University of Tennessee. PhD Dissertation, Tennessee Press, Knoxville. Department of Anthropology, University of Arizona, Tucson. Ledbetter, R. Jerald, David G. Anderson, and Scott C. Meeks Miller, D. Shane, and Joseph A. M. 2008 Recording Paleoindian Projectile Gingerich Points in Georgia. Current 2013 Regional Variation in the Terminal Research in the Pleistocene Pleistocene and Early Holocene 25:89–91. Radiocarbon Record of Eastern North America. Quaternary Loebel, Thomas J. Research 79:175–188. 2005 The Organization of Early Paleoindian Economies in the Miller, D. Shane, David G. Anderson, Western Great Lakes. PhD Thaddeus G. Bissett, and Stephen Dissertation, Department of B. Carmody Anthropology, University of Illinois, 2012 Radiocarbon Dates from Three Chicago. Sites along the Middle Cumberland River near Nashville. Tennessee McGahey, Samuel O. Archaeology 6:52–71. 2004 Mississippi Projectile Point Guide. Archaeological Report No. 31. Morrow, Juliet E. Mississippi Department of Archives 2014 Early Paleoindian Mobility and and History, Jackson, Mississippi. Watercraft: An Assessment from the Mississippi River Valley. Miller, D. Shane Midcontinental Journal of 2010 Clovis Excavations at Topper Archaeology 39:103–129. 2005–2007: Examining Site Formation Processes at an Upland Morrow, Juliet E. Paleoindian Site along the Middle 2015 Clovis Era Point Production in the Savannah River. Occasional Midcontinent. In Clovis: On the Paper No. 1 of the Southeastern Edge of a New Understanding, Paleoamerican Survey, South edited by Ashley M. Smallwood Carolina Institute of Archaeology and Thomas A. Jennings, pp. 83– and Anthropology, Columbia. 107. Texas A&M University Press, 2011 Rivers, Rocks and Eco-tones: College Station. modeling Clovis landscape-use in the Southeastern United States. Morse, Dan F., and Phyllis A. Morse Paper presented at the 76th 1983 Archaeology of the Central

144 Making a Difference

Mississippi Valley. Academic 2005 Representivity of the Midwestern Press, New York. Paleoindian Site Sample. North American Archaeologist 25:189– Prasciunas, Mary M. 212. 2011 Mapping Clovis: Projectile Points, Behavior, and Bias. American Smallwood, Ashley M. Antiquity 76(1):107-126. 2011 Clovis Technology and Settlement in the American Southeast. PhD Sain, Douglas A. Dissertation, Department of 2011 Clovis Blade Technology at the Anthropology, Texas A&M Topper Site (38AL23): Assessing University. College Station, Texas. Lithic Attribute Variation and 2012 Clovis Technology and Settlement Regional Patterns of Technological in the American Southeast Using Organization. Columbia, South Biface Analysis to Evaluate Carolina: Occasional Papers No. 2, Dispersal Models. American Southeastern Paleoamerican Antiquity 77(4):689–713. Survey South Carolina Institute of Archaeology and Anthropology, Smallwood, Ashley M., Thomas A. University of South Carolina. Jennings, David G. Anderson, and Jerald Ledbetter Sain, Douglas A. 2015 Testing for Evidence of Paleoindian 2015 Pre Clovis at Topper: Evaluating Responses to the Younger Dryas the Role of Human versus Natural Chronozone in Georgia. Agency in the Formation of Lithic Southeastern Archaeology 34:23– Deposits from a Pleistocene 45. Terrace in the American Southeast. PhD Dissertation, Department of Speth, J. D., K. Newlander, A. A. White, Anthropology, The University of A. K. Lemke, and L. E. Anderson Tennessee, Knoxville. 2013 Early Paleoindian Big-game Hunting in North America: Sassaman, Kenneth E. Provisioning or Politics?” 2010 The Eastern Archaic, Historicized. Quaternary International 285: 111– AltaMira Press, Lanham, Maryland. 139.

Sherwood, Sarah C., Boyce N. Driskell, Stanford, Dennis J., E. L. Canales, John Asa R. Randall, and Scott C. B. Broster, and Mark R. Norton Meeks 2006 Clovis Blade Manufacture: 2004 Chronology and stratigraphy at Preliminary Data from the Carson- Dust Cave, Alabama. American Conn-Short Site (40BN190), Antiquity 69:533–554. Tennessee. Current Research in the Pleistocene 23:145–147. Shott, Michael J. 2002 Sample Bias in the Distribution and Tankersley, Kenneth B. Abundance of Midwestern Fluted 2004 The concept of Clovis and the Bifaces. Midcontinental Journal of Peopling of North America. In The Archaeology 27:89–123. Settlement of the American

145 Tennessee Archaeology 8(1-2) Summer 2016

Continents: A Multidisciplinary Odyssey conference, Santa Fe, Approach to Human Biogeography, New Mexico. edited by C. M. Barton, G. A. Clark, D. R. Yesner, and G A. Pearson, Walthall, John A. pp. 49–63. Tucson: University of 1998 Rockshelters and Hunter-Gatherer Arizona Press. Adaptation to the Pleistocene/ Holocene Transition. American Thulman, David K. Antiquity 63(2):223–238. 2006 A Reconsideration of Paleoindian Social Organization in North- Waters, Michael R., and Thomas W. central Florida. PhD Dissertation, Stafford, Jr. Department of Anthropology, 2007 Redefining the Age of Clovis: Florida State University, Implications for the Peopling of the Tallahassee. Americas. Science 15(5815):1122– 1126. Thulman, David 2009 Freshwater Availability as the Waters, Michael R., and Thomas W. Constraining Factor in the Middle Stafford Paleoindian Occupation of North- 2013 The First Americans: A Review of Central Florida. Geoarchaeology the Evidence for the Late- 24(3):243–276. Pleistocene Peopling of the Americas, in Paleoamerican Walters, Sarah E. Odyssey, edited by Kelly E. Graf, 2013 Tell Us What You Ate – and We’ll Caroline V. Ketron, and Michael R. Tell You Who, What, and When Waters, pp. 541–560. Center for You Were: Paleoethnobotanical the Study of the First Americans, Recovery and Analysis along the Texas A&M University, College Savannah River. Paper presented Station. at the 79th annual meeting of the Society for American Archaeology, Waters, Michael R., Charlotte D. Pevny, Austin, Texas. and David L. Carlson 2016 Deeply Rooted: A Feasibility Study 2011 Clovis Lithic Technology: Testing the Potential for AMS Investigation of a Stratified Dating through Paleoethno- Workshop at the Gault Site, Texas. botanical Recovery Methods at the Texas A&M University Press, Topper Site (38AL23). Master’s College Station. Thesis, Department of Anthropology, University of David G. Anderson Tennessee, Knoxville. Department of Anthropology University of Tennessee Knoxville, TN Walters, Sarah E., Sean Cary von Gunter, and Albert C. Goodyear 2013 The Secret Life of Carbonized Plant Remains: A Guide to Dating in the Age of Clovis. Poster presented at the Paleoamerican

146